Ideas & society2023 · shared on Hacker News
Code trained LLMs reason better, on benchmarks that have nothing to do with code
Explore Code trained LLMs reason better, on benchmarks that have nothing to do with code, its original source and Hacker News discussions. Discover community perspectives and save it to your reading list.
GO BEYOND THE ABSTRACT
The conversation around the idea.
Original Hacker News threads. Different perspectives, questions, and things worth thinking about.
Showing up to 100 related threads. A paper may appear in a comment rather than the original submission.
Links found in the discussion
Up to 50 recent source links, in HN item order.
Ysimonw ·
Understanding GPT Tokenizers
29 points9 comments
Yog_kalu ·
Can Programming Languages Boost Each Other via Instruction Tuning? Yes
10 points1 comments
Yhhs ·
LLMs are not greater than the sum of their parts: Stanford researchers
4 points0 comments
Yanticrymactic ·
Ask HN: Why can GPT4 en/decode base64?
3 points3 comments
Ybelter ·
Chatbots Don’t Know What Stuff Isn’t
2 points0 comments
Yog_kalu ·
Code trained LLMs reason better, on benchmarks that have nothing to do with code
2 points0 comments
Ywskinner ·
AI Safety and the Age of Dislightenment
2 points0 comments
Ydnqthao ·
Language Models of Code Are Few-Shot Commonsense Learners
1 points0 comments
Ytim_sw ·
Language Models of Code Are Few-Shot Commonsense Learners
1 points0 comments