What Would Falsify It? A Variable Specific Evidence Standard for Mechanistic Claims About Self Explanation
cs.AI, cs.LG
Submitted: 2026-09-26
Updated: 2026-09-26
Terminology
Sources
- Looking Inward: Language Models Can Learn About Themselves by Introspection
- Reasoning Models Don't Always Say What They Think
- Language Models (Mostly) Know What They Know
- Measuring Faithfulness in Chain-of-Thought Reasoning
- Emergent Introspective Awareness in Large Language Models
- Is This the Subspace You Are Looking for? An Interpretability Illusion for Subspace Activation Patching
- Walk the Talk? Measuring the Faithfulness of Large Language Model Explanations
- Can LLMs Introspect? A Reality Check
- When Decodability Is Not Enough: Logical Validity Representations, Behavioral Dissociation, and Causal Tests in Language Models
- Decodability is Not Causality: Dissociating Probe Readouts from Behavioral Drivers via SAE Decomposition
- Towards Faithful Natural Language Explanations: A Study Using Activation Patching in Large Language Models
- Towards Best Practices of Activation Patching in Language Models: Metrics and Methods
Related papers
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
- MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
- Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems
- VSAL: A Vision Solver with Adaptive Layouts for Graph Property Detection