Self-Reports Do Not Identify Self-Models: An Identifiability Test for Counterfactual Reports
cs.CL
Submitted: 2026-09-26
Updated: 2026-09-26
Terminology
Sources
- Causal Abstractions of Neural Networks
- The Llama 3 Herd of Models
- Measuring Faithfulness in Chain-of-Thought Reasoning
- Training Language Models to Explain Their Own Computations
- Language Models Are Capable of Metacognitive Monitoring and Control of Their Internal Activations
- Mechanisms of Introspective Awareness
- The Geometry of Truth: Emergent Linear Structure in Large Language Model Representations of True/False Datasets
- Qwen2.5 Technical Report
- On Causal and Anticausal Learning
- Towards Understanding Sycophancy in Language Models
- Steering Language Models With Activation Engineering
- Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Prompting
- Manifold Steering Reveals the Shared Geometry of Neural Network Representation and Behavior
- Representation Engineering: A Top-Down Approach to AI Transparency
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