https://asgsaeid.github.io/blog/ Notes on evaluating and improving tool-using agents. en https://asgsaeid.github.io/blog/2026-10-your-judge-isnt-lenient-or-strict-your-rubric-is.html https://asgsaeid.github.io/blog/2026-10-your-judge-isnt-lenient-or-strict-your-rubric-is.html Fri, 16 Oct 2026 09:00:00 -0700 Holding the judge and the outputs fixed, what the rubric asks for moved the pass rate on correct outputs from 2% to 98%. Swapping the judge model barely mattered. https://asgsaeid.github.io/blog/2026-10-self-review-erases-what-you-taught-your-rubric.html https://asgsaeid.github.io/blog/2026-10-self-review-erases-what-you-taught-your-rubric.html Thu, 15 Oct 2026 09:00:00 -0700 Four rounds of model self-review pulled two good rubrics toward the reviewer’s own standard. Only delete-only revision was safe. https://asgsaeid.github.io/blog/2026-09-agents-that-design-their-own-judges.html https://asgsaeid.github.io/blog/2026-09-agents-that-design-their-own-judges.html Tue, 15 Sep 2026 09:00:00 -0700 Five findings from a matched, five-repeat benchmark of self-designed evaluators, ACE and a ten-label expert calibration on tau2-bench and AppWorld.