U-WIN/equational-theory-stage2-qwen38-graded

Equational Theories Stage II proofs, regraded by Qwen3.8-27B Saved Stage II proof attempts from Qwen3.5-9B-based models, each graded twice on the same 0–7 rubric against the same verified reference proof: the original GPT-5.6-sol grade and a regrade by Qwen3.8-27B (thinking enabled, medium reasoning effort). The models are Base, every published note/trajectory mixture scale (1M–500M, including both 500M generators) and Notes NTP with KL=0.01 at 50M, 100M and 350M. There are 139… See the full description on the dataset page: https://huggingface.co/datasets/U-WIN/equational-theory-stage2-qwen38-

<iframe src="https://silytics.si/embed/item/37fdbfbf-8772-45e1-9f8d-a5547f6df58a" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 21 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.