Atlas

Benchmarks

RSI-Infra Bench Preview (ECI-fitted score)

Peking University DAGroup's RSI-Infra Bench Preview asks whether a coding agent can make real training jobs from public repositories run faster in 10 hours without changing the training. Its eight tasks (LLaVA, YOCO, NanoChat, BAGEL SFT, R3, 3D Gaussian splatting, and Slime GRPO and on-policy distillation) run on one to four Hopper GPUs; a frozen judge times the original and the agent's code back to back and accepts only bit-identical results or results within 3 sigma of the original's seed-to-seed variation, at no lower precision, and a rejected or slower run counts as 1x. Each task's speedup is normalized as ln(speedup) / ln(best speedup on the task), and an Epoch Capabilities Index-style fit over all models and tasks gives the score, scaled so that Claude Opus 5.5 = 40 and Kimi K3.1 Pro = 10. Each model joining the board refits the scale and can move every other score, so the score is relative to the board's cohort and this is a diagnostic.

Models
9
Top score
40.0
Leader
Claude Opus 5.5
Updated
Nov 18, 2026
Resolution
—
Saturation
—

No signal or difficulty estimate. Held out of the fit by design.

Organizations
9 dated models · 2 frontier modelsClosed weightsOpen weights
Data: Peking Universityatlas.village.dev

RSI-Infra Bench Preview (ECI-fitted score) leaderboard

Loading Atlas data…