Ideas & society2025 · shared on Hacker News
The SWE-Bench Illusion: When LLMs Remember Instead of Reason
Explore The SWE-Bench Illusion: When LLMs Remember Instead of Reason, 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.
Ylackoftactics ·
Open models by OpenAI
2,124 points876 comments
Ydmitrybrant ·
Using Claude Code to modernize a 25-year-old kernel driver
929 points318 comments
Ymustaphah ·
Top model scores may be skewed by Git history leaks in SWE-bench
466 points153 comments
Ycarlotasoto ·
Principles for production AI agents
128 points19 comments
Ylogicprog ·
Did Claude Increase Bugs in Rsync?
39 points15 comments
Ynurple ·
The SWE-Bench Illusion: When LLMs Remember Instead of Reason
4 points0 comments
Ycadabrabra ·
Prediction: Claude 5 will be a major regression
3 points8 comments
Ym463 ·
FSF Threatens Anthropic over Infringed Copyright: Share Your LLMs Freel
2 points0 comments
Ycadabrabra ·
The SWE-Bench Illusion: When LLMs Remember Instead of Reason
1 points0 comments