Know the Normal, Track the Attack: Context-Grounded and Stateful LLM Investigation over System Provenance

arXiv:2609.36494 · cs.CR · Submitted 2026-09-29 · Read on arXiv

cs.CR

Submitted: 2026-09-29

Updated: 2026-09-29

Code: https://github.com/darpa-i2o/Transparent-Computing

Terminology

• Provenance — 50× in this paper · explained in 4 episodes
• Accuracy — 14× in this paper · explained in 1 episode
• Ablation — 8× in this paper · explained in 1 episode
• Semantic Gap — 6× in this paper · explained in 2 episodes
• Core — 5× in this paper · explained in 2 episodes
• Adam — 4× in this paper · explained in 2 episodes
• DARPA Transparent Computing — 4× in this paper
• Grounding — 4× in this paper · explained in 3 episodes
• Embeddings — 3× in this paper · explained in 2 episodes
• Privilege Escalation — 3× in this paper
• Benchmark — 2× in this paper · explained in 2 episodes
• Calibration — 2× in this paper · explained in 7 episodes
• Internal Reconnaissance — 2× in this paper
• Reconstruction Error — 2× in this paper · explained in 2 episodes
• Reinforcement Learning — 2× in this paper · explained in 4 episodes
• Representation Learning — 2× in this paper · explained in 2 episodes
• SIGSAC Conference on Computer and Communications Security — 2× in this paper
• Trace — 2× in this paper · explained in 2 episodes
• Adaptation — 1× in this paper · explained in 1 episode
• Adversarial Attacks — 1× in this paper · explained in 1 episode
• Alert Fatigue — 1× in this paper · explained in 1 episode
• Consistency — 1× in this paper · explained in 3 episodes
• Covering Tracks — 1× in this paper
• Feature Extraction — 1× in this paper · explained in 2 episodes
• Graph Construction — 1× in this paper · explained in 3 episodes
• Hallucination — 1× in this paper · explained in 3 episodes
• Heterogeneous Graph — 1× in this paper · explained in 2 episodes
• Initial Compromise — 1× in this paper
• LLM judge — 1× in this paper · explained in 2 episodes
• Sparsity — 1× in this paper · explained in 4 episodes

Sources

• Relational Graph Attention Networks
• DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence
• Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Training
• Prompt Injection attack against LLM-integrated Applications
• Efficient Estimation of Word Representations in Vector Space
• BadGPT: Exploring Security Vulnerabilities of ChatGPT via Backdoor Attacks to InstructGPT
• Masked Label Prediction: Unified Message Passing Model for Semi-Supervised Classification
• Graph Attention Networks
• Universal and Transferable Adversarial Attacks on Aligned Language Models

Related papers

• SoK: AI-Augmented Binary Reversing
• Relaxed Sender Anonymity for CBDC Interbank Settlement: A Zero-Knowledge Approach on Permissioned EVM
• Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages
• Efficient Fuzzy PSI under One-Sided Assumptions
• Sealing the Audit-Runtime Gap for LLM Skills
• Token Composition: A Graph Based on EVM Logs