Guide · 42 episodes · updated 2026-09-06

Agent benchmarks: how the papers measure, game, and misread them

← all guides

Why do agent benchmark scores vary so much depending on grading method and task design?

Agent benchmarks try to score whether a system can actually finish a multi-step job, like fixing a bug across a or completing a task in a real app, rather than just answering a question. The papers keep hitting the same wall: , grading method, and task design shape the number as much as the model does. State-based grading on exposes that claim success without doing the work, while public benchmarks routinely diverge from how phones behave in practice. Several episodes build purpose-built suites instead of trusting existing ones, targeting , poisoned inputs, or ambiguous instructions specifically. Others worry the benchmark and the method it evaluates were co-designed, quietly inflating results. The recurring disagreement is whether a high score reflects real or just a benchmark an agent learned to satisfy.

What agent benchmarks means

Agent benchmarks measure how well AI systems perform multi-step, tool-using tasks — navigating a browser, fixing a bug across a repo, completing a research task — rather than answering a one-shot question. They typically score end-to-end task completion, and their results are notoriously sensitive to scaffolding choices.

The episodes (42)

Newest first. Each line is what that paper contributed to the question.

Papers we have not covered yet

Other guides

Intro written by Anthropic's Claude Sonnet 5; episodes selected and edited by Garrett Casey. Episode notes come from each episode's own analysis. How PaperDive is made.