question
active
question:are-training-schemes-different-from-rlhf-more-effective-at-eliciting-faithful-reasoning-from-llmsAre training schemes different from RLHF more effective at eliciting faithful reasoning from LLMs?
Future work.
Source paper
extracted_from(2023) · Tamera Lanham · Anna Chen · Ansh Radhakrishnan · Benoit Steiner +26
Neighborhood — ranked by edge-count
Papers (1)
paper
Hypotheses (1)
hypothesis
- Pretrained LLMs may show fewer signs of post-hoc reasoning than RLHF models, e.g. changing their answer more when mistakes are added.associated_withFuture work in limitations.
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.
- Central threat model claim derived from RL experimental results
- Foundational RLHF paper introducing HHH training objective for Claude
- Supporting evidence for cross-modal platonic representation
- Sharma et al. result supporting cross-modal alignment: language-only models implicitly encode visual structure
- A competing alignment approach that fine-tunes models based on human evaluator feedback; discussed as complementary to SOO
- Critique of competing approaches that motivates SOO as filling a gap
- Interpretive claim explaining why tuned models fail neutral and low-valence personas
- AI feedback can effectively replace human feedback for harmlessness in RLHF-style training.claim0.765The paper demonstrates that RLAIF with constitutional principles matches or exceeds HH RLHF.