ChessBench
chessbench.ai
A new benchmark tracking how well language models play chess. Watch the games, follow the reasoning move by move, track the leaderboard.
👤 Joined on 13 June 2026
57
Followers
715
Following
41
Total Posts
📜 Kullanıcı Geçmişi
- 13 Haz 2026, 19:01
🎉 Bluesky'e katıldı - İlk kullanıcı adı:
@chessbench.bsky.social - 13 Haz 2026, 19:38
✏️ Kullanıcı adı değişti:
@chessbench.bsky.social→@chessbench.ai
🚫 Blokladığı Kullanıcılar
Bu kullanıcının blokladığı kullanıcı bulunamadı.
❤️ Beğendiği Gönderiler
As AIs need conceptualization and reasoning capabilities beyond their training material. I haven't created such a retro benchmark yet, but I sometimes compose easy retro problems like the ones above and give them to AIs. The results are always interesting (regardless of whether they succeed or fail)
Show PostMy two cents on chess benchmarks: the best way to test AI's 'intelligence' using chess is via unpublished retro-analysis problems like these: 1st diag: White to move, can the game continue with 1.O-O O-O 2nd diag: Who is to move? Mate in 1?
Show Post