OpenAI matrix multiplication papers spark debate: middle-school combinatorics beat CW-style bounds
basedjensen · x · 2026-10-12
Tarun Chitra's takeaways from OpenAI's matrix and integer multiplication papers: the Coppersmith-Winograd line led TCS astray into thinking wreath products would beat middle-school combinatorics, and Õ-notation was a mistake that made a generation ignore log factors.
More from Research
- AWS+HSBC paper co-evolves agent graders from just 10 labeled examples, matching ground truth — rohanpaul_ai · 2026-10-12
- Nature MI paper: LLMs discover complementary heuristics for combinatorial optimization — lulzxdxdxd · 2026-10-12
- willcb: inverse RL ideas like PRIME set for resurgence on non-verifiable tasks — willcb · 2026-10-12
- 中国顶尖数学家集体站台:支持用 AI 做数学研究 — Separate_Lock_9005 · 2026-10-12
- RSI-Anything uses world models for recursive self-improvement, lifting lab yields to 96% — jiqizhixin · 2026-10-12
- Kling's DiffusionART cuts video RL fine-tuning cost by 12x while raising rewards up to 53% — KlingTeam · 2026-10-12