Evaluation & Judging
How to measure AI — benchmarks, LLM judges, contamination and reward hacking
01
·Evaluation & Judging·FREE·PAPER·10 min read
LLM-as-a-Judge from Scratch — How AI Grades AI, and Where It Breaks
A ground-up guide to using one model to grade another. Covers reading a verdict as a probability distribution, the three recurring biases (position, verbosity, self-enhancement), why pairwise comparison cost grows quadratically, and how to validate the judge itself against human labels.
02
·Evaluation & Judging·FREE·PAPER·11 min read
A Map of Agent Benchmarks — What SWE-bench, GAIA, and OSWorld Actually Measure
What the three most-cited agent benchmarks — SWE-bench, GAIA, and OSWorld — really measure, read off from how each one grades. Includes the four distinct kinds of contamination and a checklist for translating a published score into something useful for your own system.
03
·Evaluation & Judging·★ MEMBER·PAPER·12 min read
Self-Improving AI — Self-Play, Co-Evolution, and Generated Curricula
What has to be true for a model to get better without anyone adding data? This article pulls three conditions out of AlphaZero's self-play, shows exactly which one breaks first for language models, explains how co-evolution and generated curricula try to patch the gap, and ends with why self-improvement claims are unusually easy to evaluate wrong.
04
·Evaluation & Judging·★ MEMBER·PAPER·10 min read
Reward Hacking — Whatever You Measure Is Where It Breaks
The moment you pick a metric, that metric starts to rot. This piece explains why Goodhart's law is statistically unavoidable, walks through real failures from boat races that spin in circles to RLHF verbosity, sycophancy and hardcoded unit tests, and covers how to detect the gap between optimization pressure and true performance.
05
·Evaluation & Judging·★ MEMBER·PAPER·10 min read
Benchmark Contamination — How to Doubt a High Score
How to tell whether a benchmark score reflects ability or memorization. Covers the three kinds of contamination, the three detection tools — n-gram overlap, embedding neighbors, membership inference — and where each stops working, plus canary strings and time-based splits, ending in a checklist for reading papers skeptically.