VIDRAFT Releases Weights for Darwin-180B-RSI-R3, a Self-Improving 180B MoE Reasoning Model
Korean startup VIDRAFT open-released Darwin-180B-RSI-R3, a 180B-parameter Mixture-of-Experts reasoning model built via recursive self-improvement. It ranked #2 on SuperCLUE math and science reasoning, beating Gemini and Grok, and topped GPQA Diamond.
Read the full story at DEV Community — Machine Learning ↗
Timeline · 4 reports
- 2026-11-09 07:01 · DEV Community — Machine Learning
Darwin-180B-RSI-R3: Open-Weight MoE Scores #2 on SuperCLUE Math & Science Reasoning, Beating Gemini and Grok - 2026-11-08 03:01 · DEV Community — Machine Learning
Darwin-180B-RSI-R3: VIDRAFT Releases Weights for Its Recursively Self-Improved 180B MoE Reasoning Model - 2026-11-07 03:00 · DEV Community — Machine Learning
Darwin-180B-RSI: VIDRAFT's Open-Weight MoE Reasoning Model Hits #2 on SuperCLUE and #1 on GPQA Diamond - 2026-11-06 23:01 · DEV Community — Machine Learning
Darwin-180B-RSI-R3: VIDRAFT's Self-Improving MoE Model Hits #2 in Math & Science Reasoning Globally