Latest / Logosofia no Dia a Dia
How to Spot Gamed Benchmarks in Chinese AI
With Chinese AI models flooding the market at rock-bottom prices, how can developers tell real reasoning power from memorized answers? Herman and Corn dissect data contamination, agentic evaluation, and the benchmarks that actually matter. Episode 130 — myweirdprompts.com/n/130 (permanent short link: myweirdprompts.com/403711)
The skinny
The skinny isn't ready yet — notes appear once the transcript is processed.