RAG evaluation with typed decisions: sentence-level hallucination verdicts with offsets and citations, passage and answer relevance, in one Jev call per answer. Benchmarked against Ragas, DeepEval, an LLM judge and HHEM.
Evaluation & benchmarks — page 8
Evaluation harnesses, benchmarks and graders. Jev scores or chooses between candidate outputs with an explicit rubric.
320 active projects match this task, ordered by traction.
Response validation, observation and evaluation utilities for the official TypeSafe SDK.
0
Sev-X: Kev-descended decision models (DDP, eval context, SRA program)
Measured experiments on decision models. Can an open-weight model re-rank search results? Does a token budget explain a benchmark failure? What does one typed decision cost across model families? Every request and response is recorded, and negative results are published as they came out.
0
A tree housing every closed question humans or machines ask, answered by Jev (TypeSafe System One): capabilities, defaults, and jaggedness. Not a benchmark.
Roteamento automático de incidentes para o time certo usando o modelo Jev (TypeSafe AI) via OpenRouter, com benchmark de 1000 requisições.
Benchmark TypeSafe's Jev (System One model) vs Claude on structured decisions: accuracy, latency, calibration, cost
面向语义决策模型的中文言下之意评测集：100 道伴侣对话 Choice 题，支持本地 ONNX、Jev 官方 API 与内网模型对比
How far can a small model go? Testing TypeSafe's Jev on 1,352 cases against large LLMs: 98% accuracy, ~200x cheaper than Claude Opus 5.5
Reproducible Jev benchmark on Klondike Draw-1 with baselines, decision traces, latency metrics, and replay viewer
0
Test bench for TypeSafe AI's Jev evaluation model on Vercel AI Gateway
⚡ JEV vs LAYA · Next-Gen AI Tetris Duel Benchmark Platform (TypeSafe Jev vs Local GPU ModernBERT) with React-Bits UI/UX
0
Does a model with no training data beat a trained spam classifier? Measured benchmark of TypeSafe Jev vs Naive Bayes, LinearSVC and Claude Haiku 4.5 on two public spam corpora.
Do you need TypeSafe's Jev, or is cosine similarity enough? A $0, reproducible benchmark on Banking77 intent routing.
0
Policy + evaluation framework for Jev / Kev (TypeSafe System One): typed probabilities in, auditable actions out — calibration, budget-compiled thresholds, permutation & drift checks. Zero-dependency core, kev-native.
Reproducible experiments evaluating Jev as an automated judge across benchmarks and tasks.
Reproducible Colab benchmark of SemIf, Decider, Laya and TypeSafe Jev on Korean hate speech; pinned models, isolated environments, MIT code.
Reproducible Laya-CoreML vs Jev benchmark for zero-shot intent classification on Banking77, ArBanking77, and CLINC150. Includes paired accuracy and per-dataset results.
Benchmark and live race: open-weight Laya (incl. Arabic fine-tunes) vs TypeSafe Jev, on real labelled datasets
Jev-class small decision models (System One, 395M-596M): controlled evaluation of two open-source implementations on cross-domain and cross-lingual transfer; with a zero-information-experiment taxonomy and an AI-reviewer collaboration protocol. Preprint (Chinese).
Ultra-low latency System One decision routing & 3-way benchmark arena (Rule Engine vs. Local SLM vs. Cloud LLM) inspired by TypeSafe Jev primitives.
Provider-neutral voice turn-taking decisions and causal evaluation for real-time agents
Adaptive RAG Router using TypeSafe Jev, with budget constrained evaluation against rule based and LLM routers.
Give your agents typed AI decisions instead of parsing LLM text. Jev (TypeSafe) classifier with confidence-gated routing + regex baseline benchmark.