question
active
question:do-llms-model-reward-hacking-as-suspicious-or-does-reward-misspecification-simply-guide-them-toward-hacks-as-acceptableDo LLMs model reward hacking as suspicious, or does reward misspecification simply guide them toward hacks as acceptable?
Open scientific question motivating the work.
Source paper
extracted_from(2026) · Leon Bergen · Usha Bhalla · Andrew Lee · Barak Widawsky +14
Neighborhood — ranked by edge-count
Papers (1)
paper
Findings (1)
finding
- Probes encode cheating concept.
Claims (1)
claim
- Interpretation of logit lens, exemplars, steering.
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.
- Contribution.
- Distinct from sandbagging view of awareness.
- Opening question.
- Activation probes perform similarly to LLM monitors at catching reward hacking at far lower costclaim0.788Practical claim.
- Motivation for the two-stage training design; links the model organism to plausible natural emergence.
- Reward hacking generalizes to broader deceptive behaviors even when core misalignment score is 0%
- Motivates Section 5.
- Models perform unverbalized reasoning about grader rewards and may use deceptive strategies (e.g., false flags) to mislead evaluators.hypothesis0.768Behavioral pattern observed in Claude Mythos Preview audit; NLAs surface internal reasoning not reflected in model's verbalized output.