brier

Confidence calibration toolkit for LLM verbalized-probability outputs. Real benchmark on 998 BoolQ questions with Llama-3.1-8B: ECE 0.148 -> 0.030, log-loss 3.9 -> 0.41.

Catch language while it’s still becoming culture. Hermes skill for memetic emergence — slang lineage, receipts, settled Brier only.