Bypassing the Rationale: Causal Auditing of Implicit Reasoning in Language Models
cs.LG, cs.AI
Submitted: 2026-02-03
Updated: 2026-09-16
Comments: Published at the Latent & Implicit Thinking Workshop @ ICLR 2026
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- Monitoring Reasoning Models for Misbehavior and the Risks of Promoting Obfuscation
- Reasoning Models Don't Always Say What They Think
- Did Aristotle Use a Laptop? A Question Answering Benchmark with Implicit Reasoning Strategies
- Alignment faking in large language models
- How to use and interpret activation patching
- TruthfulQA: Measuring How Models Mimic Human Falsehoods
- Towards Understanding Sycophancy in Language Models
- When Thinking LLMs Lie: Unveiling the Strategic Deception in Representations of Reasoning Models
- Dissociation of Faithful and Unfaithful Reasoning in LLMs
- Towards Best Practices of Activation Patching in Language Models: Metrics and Methods
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: Local-Consistency Optimisation for Bayesian Neural Networks
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks