claim
active
claim:faithful-cots-depend-on-prior-knowledge-of-the-answer-which-may-explain-lower-faithfulness-on-harder-questionsFaithful CoTs depend on prior knowledge of the answer, which may explain lower faithfulness on harder questions
Mechanism explanation.
Source paper
extracted_from(2025) · Yanda Chen · Joe Benton · Ansh Radhakrishnan · Jonathan Uesato +11
Neighborhood — ranked by edge-count
Papers (1)
paper
Claims (1)
claim
- Lower faithfulness on GPQA.
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.
- Framing of defense-in-depth.
- For reasoning faithfulness, larger models may behave worse than smaller ones because they can predict answers without CoT.hypothesis0.856Motivates §3.
- From decoupling of AOC and accuracy gain.
- it seems hard to get an exact estimate of CoT faithfulness without solving interpretabilityquote0.826Limit on measurement.
- Generality of methods.
- Interpretation of §3.
- Whether stated reasoning accurately represents the reasoning process behind the prediction.
- In motivated reasoning the CoT derives from the human hint, but not in faithful reasoning.hypothesis0.814Tested by varying the suggested answer.