Researchers Estimate GPT, Claude, and Gemini Parameter Counts Using API Calls

有人只用API就猜出了GPT、Claude、Gemini的参数量？社区吵翻了

Bojie Li posted IKP on arXiv to estimate parameter counts of 188 LLMs from 27 vendors via black-box API calls. The dataset has 1,400 questions across 7 rarity tiers, fitted on 89 open models with R²=0.917. Debate centers on synthetic data, MoE effects, and a 90% interval of 0.3x to 3x.

Why it matters: HKR-H/K/R all pass: API-only parameter inference is a strong hook, with concrete counts and error bounds. The 0.3–3x CI limits confidence, so this fits 78–84 featured, not P1.