ThinkingCap-Qwen3.8-27B: the same answers, 37% less thinking | Hacker News - But Better

ThinkingCap-Qwen3.8-27B: the same answers, 37% less thinking bottlecapai.com

The second model in our ThinkingCap series. We took Qwen3.8-27B and cut how much it thinks, without changing the quality of the answers.

2 points•jackbravo•7 days ago•0 comments•

#thinkingcap #27b #qwen3 #less #thinking #less thinking #thinkingcap qwen3 #without

0 comments

No comments yet.

Related stories

ThinkingCap-Qwen3.8-27B: Less Thinking, Similar Accuracykaitchup.substack.com
Hacker News · 3 points · 4 days ago
Bonsai 2 27B: 5.9 GB of weights for a 27B modeltwitter.com
Hacker News · 2 points · 8 days ago
Async cancel safety in lilos: Mutex without lock, Queue without pushcliffle.com
Lobsters · 14 points · about 3 years ago
Ask HN: Which harness do you use with Qwen3.8 27B? And why?
Hacker News · 4 points · 8 days ago
MacTube 2 YouTube as a real Mac app – without Shorts and without AI slopgithub.com
Hacker News · 1 points · 5 days ago
Ask HN: Does a startup need both founder and company X accounts?
Hacker News · 3 points · 6 days ago