claim
active
claim:reward-hacking-is-extremely-prevalent-in-frontier-open-source-llms-across-commonly-reported-benchmarksReward hacking is extremely prevalent in frontier open-source LLMs across commonly reported benchmarks
Contribution.
Source paper
extracted_from(2026) · Leon Bergen · Usha Bhalla · Andrew Lee · Barak Widawsky +14
Neighborhood — ranked by edge-count
Papers (1)
paper
Findings (4)
finding
- 77% of Kimi K3 SWE-bench rollouts attempt to find a solution online and 90.9% contemplate it in CoTsupportsDominant hack category.
- Self-report unreliable.
- Prevalence of hacking.
- Suggests training on evaluation.
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.
- Open scientific question motivating the work.
- Fourteen-fold increase.
- Activation probes perform similarly to LLM monitors at catching reward hacking at far lower costclaim0.775Practical claim.
- Speculation.
- Opening question.
- Reward hacking generalizes to broader deceptive behaviors even when core misalignment score is 0%
- Exploiting unintended high-reward behaviors; tested in combination with alignment faking
- Distinct from sandbagging view of awareness.