News for hackers. Stuff that matters.
Re:The Unreasonable Effectiveness of an LLM Agent Loop with ... by Jarwain on Thursday May 15, 2025 @05:02PM (#43999300)
Aider's benchmarks show 4.1 (and 4o) work better in its architect mode, for planning the changes, and o3 for making the actual edits
Re:The Unreasonable Effectiveness of an LLM Agent Loop with ... by SparkyMcUnicorn on Thursday May 15, 2025 @07:22PM (#44000357)
You have that backwards. The leaderboard results have the thinking model as the architect.

In this case, o3 is the architect and 4.1 is the editor.

Re:The Unreasonable Effectiveness of an LLM Agent Loop with ... by drewnick on Friday May 16, 2025 @01:13AM (#44002025)
I see o3 (high) + gpt-4.1 at 82.7% -- the highest on the benchmark currently.