Weco research shows autonomous AI agent achieves recursive self-improvement
The agent optimized its own codebase to beat human benchmarks.
Combined views
420.5K
17 Sources, first seen ago
Sources
ZW 
Zhaoran Wang@zhaoran_wang- likes: 2
- replies: 0
- bookmarks: 1
- reposts: 0
MN Minh Nhat Nguyen@menhguinjensen huang rn: https://twitter.com/zhengyaojiang/status/2077079778793042425
- likes: 175
- replies: 2
- bookmarks: 18
- reposts: 1
BT Bojan Tunguz@tunguzAnd here we go. https://twitter.com/zhengyaojiang/status/2077079778793042425
- likes: 600
- replies: 5
- bookmarks: 398
- reposts: 30
MJ Minqi Jiang@MinqiJiangI love this work from @zhengyaojiang, @DhruvSrikanth, @yuxiangwu_, @BingchenZhao, and the @WecoAI team. It cogently paints the path to RSI. Now the track is marked, may the race to ignition and beyond begin. https://twitter.com/zhengyaojiang/status/2077079778793042425
- likes: 27
- replies: 1
- bookmarks: 12
- reposts: 5
YO Yohei@yoheinakajimasome good self improvement research here https://twitter.com/zhengyaojiang/status/2077079778793042425
- likes: 54
- replies: 0
- bookmarks: 26
- reposts: 6
ET Ethan@torchcompiledthere is arguably a new age of "symbolic training" that is powerful on its own, and moreso in tandem with regular gradient training, where an agent performs search in optimizing its own context and harness. https://twitter.com/zhengyaojiang/status/2077079778793042425
- likes: 78
- replies: 4
- bookmarks: 29
- reposts: 4