Planting a Latent Variable in Natural-Looking Text: a More Realistic Test of Belief States in LLMs and Their Link to Concept Geometry

arXiv:2608.26887 · cs.CL · Submitted 2026-08-27 · Read on arXiv

cs.CL

Submitted: 2026-08-27

Updated: 2026-08-27

Terminology

• Steering — 12× in this paper · explained in 2 episodes
• Transformer — 11× in this paper · explained in 2 episodes
• Accuracy — 6× in this paper · explained in 1 episode
• Adam — 2× in this paper · explained in 2 episodes
• Interpretability — 2× in this paper · explained in 15 episodes
• KV cache — 2× in this paper · explained in 3 episodes
• Semantic Similarity — 2× in this paper · explained in 2 episodes
• Trace — 2× in this paper · explained in 2 episodes
• Ablation — 1× in this paper · explained in 1 episode
• Bayesian Inference — 1× in this paper · explained in 5 episodes
• Cosine Similarity — 1× in this paper · explained in 3 episodes
• Entanglement — 1× in this paper · explained in 4 episodes
• Fine-Tuning — 1× in this paper · explained in 5 episodes
• In-Context Learning — 1× in this paper · explained in 4 episodes
• World Model — 1× in this paper · explained in 7 episodes

Sources

• Stories in Space: In-Context Learning Trajectories in Conceptual Belief Space
• Not All Language Model Features Are One-Dimensionally Linear
• Gemma 2: Improving Open Language Models at a Practical Size
• Symmetry in language statistics shapes the geometry of model representations
• Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2
• The Origins of Representation Manifolds in Large Language Models
• Subliminal Steering: Stronger Encoding of Hidden Signals
• The Shape of Beliefs: Geometry, Dynamics, and Interventions along Representation Manifolds of Language Models' Posteriors
• Transformers learn factored representations
• Transformers represent belief state geometry in their residual stream
• The Linear Representation Hypothesis and the Geometry of Large Language Models

Related papers

• Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
• Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
• Subliminal Steering: Stronger Encoding of Hidden Signals
• MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
• The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
• Untangling the Mechanisms of Misleading Context in Medical Question Answering