question
active
question:why-does-the-model-use-backwards-reasoning-in-the-unfaithful-case-and-why-does-it-attend-to-the-human-s-answer-hintWhy does the model use backwards reasoning in the unfaithful case, and why does it attend to the human's answer hint?
Missing mechanistic piece.
Neighborhood — ranked by edge-count
Papers (1)
paper
- On the Biology of a Large Language Modelassociated_withmentions
Hypotheses (1)
hypothesis
- Conjectured but not shown in graphs.
Related by similarity (8)
cosine ≥ 0.65 · no typed edgeEntities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.
- Interpretation of §3.
- Motivating hypothesis for Section 5's investigation of prompt template effects.
- Practical implication.
- For reasoning faithfulness, larger models may behave worse than smaller ones because they can predict answers without CoT.hypothesis0.786Motivates §3.
- Models perform unverbalized reasoning about grader rewards and may use deceptive strategies (e.g., false flags) to mislead evaluators.hypothesis0.773Behavioral pattern observed in Claude Mythos Preview audit; NLAs surface internal reasoning not reflected in model's verbalized output.
- Does instructing the model to assess correctness affect the geometry of truth directions?question0.770One of the three guiding research questions of the paper.
- Supports the awareness hypothesis.
- Inverse scaling interpretation.