News for hackers. Stuff that matters.

A comment on Computer Anthology: A continuously evolving benchmark family for AI agents.

Re:Computer Anthology: A continuously evolving benchmark fam... by kmiens on Tuesday August 04, 2026 @10:52AM (#49169816)
The comparison between harnesses is very nice. Interesting to see that using a different harness can bump the performance of the model as much as a new version (e.g., GPT 5.5+Codex ~= GPT 5.6+Terminus, at lower cost)
[ Parent | on HN ]
Re:Computer Anthology: A continuously evolving benchmark fam... by dvaplima on Tuesday August 04, 2026 @01:10PM (#49171799)
That’s a nice discussion. Some people say that with current model capabilities, the real differentiator is the harness. What are the best harnesses you guys are using?
[ Parent | on HN ]