タグ：knowledge-distillation

Distill a repeated Jev classification task into a local model, on the fly — same answers, your hardware.

Jev-style typed decisions (yes/no, choice, score with calibrated probabilities) on HRM-Text-1B, with benchmarks against OpenJev and teacher distillation

kblam: the knowledge base for LLM-assisted mereology. it's a flat-file research/knowledge storage system with enforced organization (using heuristics, semantic similarity, and Jev evaluation) to prevent document drift, internal contradictions, incorrect data gleaned from reading superseded sections out of context, and other LLM bullshit. not done