2026-11-02 · America/Los_Angeles · 社区动态 · #20

MIT researchers say GPT-6 Astra improved a small Qwen model’s math score from 7.9% to 48.1% by directly editing just 0.016% of its weights, without gradient-based training.

MIT researchers say GPT-6 Astra improved a small Qwen model’s math score from 7.9% to 48.1% by directly editing just 0.016% of its weights, without gradient-based training.

热度 44.0 / 100；排名与评分保留该期记录。