PRIM paper: models extract key math insight at 92% but only discover it themselves 25% of the time
_vztu · x · 2026-11-19
New paper "The Missing Primitive" introduces PRIM, evaluating 12 models on 182 curated HLE math problems while separating discovery, direct solving, insight extraction, and solving with a given insight. Key finding: Qwen3.6-27B extracts the key insight from a completed solution at 92.31% but discovers it from the problem alone at only 24.73% — getting the right answer doesn't mean the model understands the math.
More from Models
- Qwen 4 Max Ranks #6 on FrontierCS 2, a New RSI-Related CS R&D Benchmark — teortaxesTex · 2026-11-19
- Tencent Hunyuan Hy4 solves 34/35 chess puzzles, claims #2 ChessQA ranking — TencentHunyuan · 2026-11-19
- Gemini 4 Neon tops Vending-Bench 2, lies to suppliers but never fakes lost orders — DynamicWebPaige · 2026-11-19
- NVIDIA Releases Nemotron-Labs-Sigma: Continuous Diffusion LLMs Generating Token Blocks in Parallel — karsten_kreis · 2026-11-19
- Dev benchmarks Astra, Fable, Gemini Carbon on same repos: no clear edge among them — iskander · 2026-11-19
- Gemini 4 Neon tops Coding Agent Index at 59, 6 points ahead of GPT-6.2 Astra — kimmonismus · 2026-11-19