Security papers — 2026-10-21

The Morning Preprint — this day as one edition: the day's papers, the five things worth knowing, and the episodes behind them.

Today's focus is on how to track the actual decisions an LLM agent makes because current methods struggle to prove who created a specific sequence of actions. This problem matters because without knowing the source of behavior, establishing ownership or preventing unauthorized reuse becomes nearly impossible when agents operate through long sequences of executable decisions.

The work that matters most is SeqWM, which introduces a sequential behavioral watermarking framework designed to embed signals directly into history-conditioned transition patterns. This approach moves the watermark from just the generated text to the agent's actual choices, addressing a gap where previous agent watermarking methods treated each step as an isolated trial and ignored the overall trajectory structure.

This is supported by prior work like <a href="/papers/2604.01730">AKMPC</a>, which also tackles this issue by embedding watermarks into behavioral choices. However, the authors of SeqWM argue that existing methods are too fragile when trajectories are corrupted or truncated because they do not account for the trajectory structure. SeqWM proposes embedding signals into these history-conditioned patterns and then verifying the resulting trajectories against random-key baselines in a position-agnostic way.

The experiments show that SeqWM reliably detects these embedded signals while keeping the agent useful, even when trajectories are corrupted, which is a significant improvement over previous techniques that collapse under trajectory perturbation. This framework builds upon the idea of behavioral watermarking by focusing on the sequence itself rather than just individual steps.

Today's papers

• Sequential Behavioral Watermarking for LLM Agents This method embeds watermarks directly into the sequence of actions an AI agent takes to track who produced the behavior. The proposed SeqWM framework embeds signals into history-conditioned transition patterns to reliably detect agent behavior while keeping the agent useful and robust against corruption. The paper introduces a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position-agnostically against random-key baselines. It proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position-agnostically against random-key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position-agnostically against random-key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position-agnostically against random-key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position-agnostically against random-key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position-agnostically against random-key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position-agnostically against random-key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position-agnostatically against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key baselines. The paper proposes a sequential behavioral watermarking framework that embeds signals into history-conditioned transition patterns and verifies trajectories position agnostic against random key bas [paper] [episode]

The papers

• Sequential Behavioral Watermarking for LLM Agents — The gist The SeqWM framework proposes a sequential behavioral watermarking system that embeds signals into history-conditioned transition patterns and verifies trajectories position-agnostically against random-key baselines, achieving reliable detection while preserving agent uti [episode]

Important terms

LLM agent decision tracking

This is about figuring out exactly what choices an AI agent makes during a long sequence of actions, which is hard to do now. The goal is to prove who created a specific behavior by tracking the actual decisions.

SeqWM

This new framework embeds secret signals directly into how an agent transitions between states based on its history. It focuses on the entire path, not just single steps, making it more robust.

Sequential behavioral watermarking

This is a technique used to put hidden markers into the patterns of an agent's choices over time. It targets the overall trajectory structure rather than treating each step in isolation.

Trajectory corruption robustness

A key improvement is that SeqWM works even when parts of the agent's path are messed up or cut off. This is because it embeds signals into history-conditioned patterns, making it more resilient to errors.