Tuesday, October 27, 2026i10X News · Section desk9 stories
Studies
Benchmarks & Evals
New benchmarks, evaluation methods and leaderboard results
FilterClaude
Filtered: Claude in Benchmarks & Evals · 6 storiesReset filter ×
-
Tampered ML: Even the Best AI Models Miss One in Five Planted Research Flaws
A new EMNLP 2026 benchmark pairs 1,858 ML experiment designs with flawed twins. Claude Opus 5 leads at 77.9%, but models flip verdicts when the order changes and favor polished flaws.
By
Editor's picks
- Tampered ML: Even the Best AI Models Miss One in Five Planted Research Flaws
- Only One AI Agent Beats Researchers’ Own Code on Most FrontierCS 2 Tasks
- Hybrid Computer-Use Agents Top Out at 41% When GUI and CLI Must Work Together
- Scientific AI Agents Finish Only 21% of End-to-End Workflows — and Often Claim Success Anyway
- AI Agents Guess Wrong at Critical Forks 40% of the Time, Taste-Bench Shows
More stories
-
Best Video Model Scores Just 51% on Visual Reasoning Tasks, VGI-Bench Finds
VGI-Bench’s 27 tasks and 810 instances show video generators can do some visual reasoning, but Seedance 2.0 tops out at 51%, with almost no self-correction during…
· RSS