Trade Offer Meme: Unsaturated RL, a Better TB 2.1 Checkpoint, and Angry Colleagues
cong_ml · x · 2026-10-22
转发者 @louvishh shares a Trade Offer meme: RL showed no signs of saturation, but colleagues declined the deal two days before release — 'I receive: curses for increasing workload; you receive: a checkpoint with better perf on tb 2.1.'
The post hints the team keeps pushing RL gains on Terminal Bench 2.1, with RL far from saturation.
More from Fun
- 'She's a 10 but that was supposed to be a random 384-bit number' — IgorBrigadir · 2026-10-22
- jxnl's One-Liner: Your Job Looks Easy for Composer to Predict — jxnlco · 2026-10-22
- Puppy Calmly Holding an Uno Card Against Four +4s Becomes the Internet's Favorite Mood — gnukeith · 2026-10-22
- Mocking the vibe-coding teen hire fantasy: 17-year-olds recruited by OpenAI from small-town China — _vztu · 2026-10-22
- Ilya Sutskever: the current AI moment feels like late-90s EFnet channels — ilyasu · 2026-10-22
- Atlas full fly-through demo stuns viewers: "the matrix might be real" — XRarchitect · 2026-10-22