Loading…

Tools

Friends

Community

A Webioinfo project · directory data from multiple sources · live metrics from the GitHub API · benchmark observations located via Benchmark Radar

RE-Bench

METR/RE-Bench

Stale

No description fetched yet.

Overview

RE-Bench is a benchmarks AI agent developed by METR with 167 GitHub stars, written in Python, currently Stale.

GitHub metrics refresh daily — last fetched today.

Hjalmar Wijk, T. Lin, Joel Becker, Sami Jawhar, Neev Parikh, Thomas Broadley … Ben West, Elizabeth Barnes

No verified run reviews yet. Be the first to run RE-Bench and report.

No reviews yet. Reviews from researchers who actually ran RE-Bench are what this platform is for.