Tag: evaluation

Frontier models scraped every word experts ever published, and it still is not enough: the judgment that makes someone worth asking was never written down. Built from an expert's consented corpus, our onixes already represent a specific human more faithfully than the strongest frontier models. The expert in the loop is the moat no scrape can reach.

Every major AI lab publishes benchmarks for reasoning, coding, and math. But nobody is measuring whether AI actually gets human expertise right. We built an evaluation framework to find out.