typed_evals
TrustifAI/typed_evals
Fast, typed, calibrated evaluations for LLM and agent outputs, powered by Jev — with simple, framework-agnostic Python APIs
- スター
- 19
- フォーク
- 1
掲載理由
Typed EvalsはJevで生成回答と記録済みエージェント実行を評価し、実行前のツール呼び出しも確認できる。
TrustifAI/typed_evals
Fast, typed, calibrated evaluations for LLM and agent outputs, powered by Jev — with simple, framework-agnostic Python APIs
Typed EvalsはJevで生成回答と記録済みエージェント実行を評価し、実行前のツール呼び出しも確認できる。
Hands-on research lab for TypeSafe's Jev (System One model): reproducible benchmarks of Noul/Choice/Score primitives, confidence gating, fan-out latency, agent control — plus a living audit of the Jev ecosystem.
共通トピック: ai-agents, evals
A pre-registered field trial of Jev (TypeSafe's judgment model) on a second brain and Claude Code history: 20 tests, bars written first, failures included, and the tools to repeat it.
共通トピック: ai-agents, evaluation
Drop-in /v1/systemone proxy that shadows Jev with local decision models (Kev, Laya) and reports whether you can swap
共通トピック: ai-agents, evaluation