Research & evaluation
Find research and evaluation projects exploring decision models and Jev-related approaches.
- Pre-registered independent eval of TypeSafe Jev against a nano-class LLM, a frontier LLM, and a supervised encoder (Banking77 + CLINC150 zero-shot)
ickma2311/jev-baselines-eval · Python
60intent-classificationllm-cascadellm-evaluationpreregistrationpython - A show-and-tell capability study for Jev, TypeSafe's System One decision model.
lbotinelly/jev-little-airways · HTML
61html - 594JevRankerJev-style decision models as BlitzRank's compare oracle: k passages scored in one parallel forward pass, zero decoded tokens — 22x faster per match than a generative listwise LLM, at 2.9x its nDCG@10.
hgqimo/JevRanker · Python
60python - 601sysone-benchFirst independent head-to-head benchmark of System One decision models (Laya vs Jev) on byte-identical inputs
instax-dutta/sysone-bench · Python
63benchmarkcalibrationdecision-modelevaluationlayallm-eval - Experimental multi-horizon BTC signal generator using TypeSafe Jev probabilities and Binance market data.
WebGrga/btc-jev-signal · TypeScript
51typescript - Learn decision models by building with them. TypeSafe Jev and Microsoft Decision-1 route support tickets in under a second for about $0.02 per 1,000 calls; Claude handles the hard cases. Interactive site, benchmark, calibration tests and docs.
shashikantdev3/decision-models-lab · TypeScript
50ai-agentsanthropicbenchmarkcalibrationclaudedecision-modelhuman-in-the-loop - 620jev-2048An instrumented 2048 web lab where every move is a Jev (TypeSafe AI System One) Choice, with no heuristic fallback | 用 Jev 决策模型驱动每一步的 2048 网页实验台,概率、置信度、延迟与成本全部摊开可见,且刻意不做启发式兜底
ARCJ137442/jev-2048 · TypeScript
512048llm-evaluationtyped-decisionstypescript - A Quick Typesafe's Jev Evaluation on Persian
ArmanJR/Jev-Persian-Benchmark · Python
50python - Daily research monitor for standing questions: deterministic Python owns the loop, TypeSafe Jev screens sources per question, Qwen writes the notes (pilot)
shimo4228/jev-research-pipeline · Python
51literature-monitoringobsidianopentelemetrypydantic-aipythonqwenresearch-pipeline - 634jev-rlJEV Reinforcement Learning: four classic games trained with JEV-powered rewards, reproducible experiments and checkpoint replays.
Bring-AI/jev-rl · Python
51python - 637jev-testA test harness for TypeSafe's Jev :Confidence audit, Hallucination judge, Jev for RAG
mity-prodgen/jev-test · Python
51evaluationmodelragpython - Fifty real-world financial use cases for TypeSafe's Jev model: typed, structured LLM answers over ledgers, fraud, portfolios, trades and filings, each graded against data where the right answer is known.
IslamBaraka90/jev-typesafe-real-financial-use-cases · JavaScript
51ai-agentsbacktestingfinancial-datafintechjavascriptllmllm-evaluation - 641jev-writerFind out which qualities of your writing actually predict engagement. Rates every post you have published against a pre-registered rubric using Jev's calibrated judgments, then tests those ratings against your real engagement numbers. Refuses to report findings your sample cannot support.
Kaos599/jev-writer · JavaScript
50agent-skillsai-gatewaycalibrated-probabilitiesclaude-codeclaude-skillcontent-analyticscreator-tools - 645jevossEvaluation tools for Jev-compatible decision models, covering accuracy, calibration, hidden instructions, padding and consistency.
mertkayacs/jevoss · Python
51benchmarkcalibrationdecision-modelevaluationllm-evaluationprompt-injectionpython - 652qwen-rlcdJev-style calibrated decision model (Choice/Score/Noul) on Qwen3.5-0.8B
shamazharikh/qwen-rlcd · Python
51python - Benchmarking Jev (Typesafe.ai) against a strong LLM on the Who&When Pro agent-failure-attribution benchmark (text subset).
TokenTrim/jev-agent-failure-benchmark · Python
40python - 702jev-playsjev-plays: TypeSafe Jev ecosystem repository.
mansicer/jev-plays · Python
40python - Zero-shot spam filtering with TypeSafe Jev Noul questions, compared with TF-IDF baselines
bitnovus/jev-spam-eval · Jupyter Notebook
41jupyter-notebook - 714Jev-VisionOpen-weight step verifier for computer-use agents: calibrated ground/skip/effect/done judgments from screenshots in ~160 ms, plus a benchmark with environment-derived labels
sseanliu/Jev-Vision · Python
40python - Three measured experiments on RAG hallucination: quote-checking, TypeSafe's Jev, and IBM's STAIR. 850+ graded questions, raw responses included.
aryanchauhanoffical/no-hallucination · Python
40benchmarkevaluationhallucinationllmragretrieval - A playground for experiments around Jev, TypeSafe's System One model.
markjaquith/typesafe-ai-playground · Rust
42rust - A playground for TypeSafeAI's Jev Model
DeepBlueDynamics/typesafe-arena · Rust
40rust - 756agentic-jev⚡ AgenticJev · 拾意 — 推荐智能体实验室。搜索 × 推荐 × 广告,Jev / MiniCPM / GPT / Claude 多模型决策与生成式组合对照。
FBddcz/agentic-jev · TypeScript
30agentic-aigenerative-recommendationminicpmrecommender-systemtypescript - 774gr-jevCode for 'When Multiple Choices Are Correct: Asymmetric Probability Allocation in Jev and Other Decision Models' (arXiv:2610.29217)
Rushab1/gr-jev · Python
30arxivcalibrationchaosnlidecision-modelecthrhotpotqallm-evaluation - 778JEVJEV 调研报告:TypeSafe System One 决策模型与开源对标(含原始核实记录)
Chorylee7/JEV · Python
30decision-modelllm-routingresearch-report - A reproducible, self-hosted benchmark for comparing JEV and LLM decision-making on your own datasets, models, and evaluation policies.
Tenkei/jev-decision-bench · Python
30ai-agentsanthropicbenchmarkdeepseekevaluationllmopenai - Jev (TypeSafe) exploratory thread: claim audit, live demos, and runnable code
SamuelSacco/jev-exploration · Python
30python - Fault Diagnosis PoC replays real compressor telemetry through an emulated Modbus device, spots suspicious behaviour with plain code, and asks a decision model which fault from the machine manual explains it.
meddle-connect/jev-fault-diagnosis-poc · TypeScript
34typescript - 799jev-freeformAn observable raw-character chat experiment powered entirely by TypeSafe Jev Choice
kesku/jev-freeform · JavaScript
30javascript - Can a System One model (TypeSafe Jev, open-source Kev) replace an LLM-as-judge in agent guardrail callbacks?
dfranco-projects/jev-guardbench · Python
30python - Comparing LLM-as-judge vs TypeSafe Jev for agent guardrails: same rules, same agent, measured on cost, latency, calibration and coverage.
deepansh-saxena/jev-guardrails · Python
31python - Putting TypeSafe's Jev to work on robots, fleets, and edge hardware — real measured numbers, honestly caveated.
robokrunch/jev-physical-ai · Python
30benchmarkedge-aiphysical-airoboticspython - Free English RAG benchmark for TypeSafe Jev 1.13 — frozen candidate pools, calibration, paired bootstrap CIs, $0 runs. Compared with OpenJev and NVIDIA cross-encoders.
emretheus/jev-rag-benchmark · Python
30benchmarkcalibrationinformation-retrievalllm-evaluationnvidiaopenjevrag - Reproducible Jev Ultrafast research-browser eval harness + field note (QC’d cases, suite runner, report generator). Not investment advice.
jgridifier/jev-research-eval · HTML
30html - Benchmarking TypeSafe's Jev decision model as a cost-efficient LLM router on RouterArena
TokenTrim/jev-routing-experiment · Python
32python - Blind security benchmarks for Jev, TypeSafe's System One model: prompt injection and vulnerable code detection, built on jev-go
Gaurav-Gosain/jev-sec-bench · Go
30go - Tool search for LLM agents: BM25 vs embeddings vs rerankers vs Jev on 525 real MCP tools, plus an experimental Jev search engine
kachar/jev-tool-search · TypeScript
30ai-agentsai-sdkbenchmarkbm25llmmcptool-search - 839jevalyzerGrade the agent sessions already on your disk. Claude Code, Codex, opencode, Gemini CLI and Antigravity, scored with Jev for cents.
killerz3/jevalyzer · TypeScript
30ai-agentsbunclaude-codeclicodexgemini-clillm-evaluation - Crypto trading bot on Binance testnet using TypeSafe (Jev) to judge news
Spykoninho/trading-bot-jev · TypeScript
30
51–100 of 389