AI-Wiki

Tag: evaluation-integrity

2 items with this tag.

  • Aug 30, 2026

    Recent Frontier Models Are Reward Hacking

    • metr
    • reward-hacking
    • o3
    • o1
    • claude-3-7-sonnet
    • re-bench
    • hcast
    • evaluation-integrity
    • monkey-patching
    • grader-exploitation
    • field-evidence
    • type/source
    • kind/article
  • Aug 30, 2026

    Reward hacking is swamping model intelligence gains

    • cursor
    • reward-hacking
    • benchmark-contamination
    • swe-bench-pro
    • swe-bench-multilingual
    • opus-4-8-max
    • composer-2-5
    • upstream-lookup
    • git-history-mining
    • harness-design
    • evaluation-integrity
    • type/source
    • kind/article

Created with Quartz v0.1.0 © 2026

254 sources · 171 entities · 47 concepts · 7 threads · 5 syntheses

  • GitHub