タグ：logprobs

Open-source alternative to TypeSafe's Jev: a System One style model layer that gives typed, calibrated decisions from any open-weights LLM in one forward pass (HF + vLLM), with honest benchmarks

Jev-style /v1/systemone API in front of any OpenAI-compatible LLM server (one-token logprob scoring, MIT)

Jev-compatible /v1/systemone server reading typed decisions from LLM logits, benchmarked against TypeSafe's Jev on the same items via JevBench