Unknown Darwin-180B-RSI posts perfect 100 on MathArena HMMT leaderboard, raising eyebrows
tensorqt · x · 2026-11-13
A MathArena/hmmtfeb2026 leaderboard screenshot is circulating on X: the top spot goes to an obscure model, FINAL-Bench/Darwin-180B-RSI, with a perfect starred score of 100 — well above Intern-S2-Preview-397B (91.57) and Thinking Machines' Inkling-Small (90.2). Known models like Qwen3.5-397B, Ling-3.0-flash, Step-3.5-Flash and NVIDIA Nemotron trail at positions 4-13.
The poster's only comment, "one is not like the others," hints the leader's provenance or score may be suspect — a notable case of unknown models topping benchmarks.
More from Models
- OpenAI's 1147-page Yang-Mills proof faces its first expert assessment: not proven yet — mishig25 · 2026-11-13
- Terminal-Bench 4.0 Cost Math: GPT-6.1 Sol Matches Opus 5.5 at ~1/8 the Price — LeadershipNo2904 · 2026-11-13
- HF staffer spotlights basal-2.0-mini, a 1.5B decision-tuned model from Bielik — lhoestq · 2026-11-13
- Observer surprised OpenAI shipped no answer to rival's new model release — jdjohnson · 2026-11-13
- Google's Carbon-Argon-Hydrogen model naming breaks its own size logic — JeremyNguyenPhD · 2026-11-13
- Claude outage: login auth issues hit claude.ai and Claude Code, plus M365 connector errors — op7418 · 2026-11-13