pull down to refresh
0 sats \ 2 replies \ @optimism 16 Oct \ on: Tencent researchers just killed fine-tuning AND reinforcement learning AI
I saw this over the weekend, still on the todo to ingest it. Here's the paper: https://arxiv.org/abs/2510.08191
My default position is skeptical. The incentives are aligned for everyone to make big claims, but underneath it is usually a trick that makes the innovation not actually significant.
Looking forward to your take though.