Research & evaluation
Find research and evaluation projects exploring decision models and Jev-related approaches.
- 1339jev-measuredMeasured cost, latency, accuracy and raw output: TypeSafe's Jev and 12 other decision models (OpenAI Decisions API, Clef, Decider, d1…) — reproducible
WallerChen/jev-measured · Python
10benchmarkdecision-modelllm-routingopenai-decisions-apipython - 1344jev-pdp-flowjev-pdp-flow: TypeSafe Jev ecosystem repository.
arlequine/jev-pdp-flow · TypeScript
11typescript - A small reproducible MuJoCo pilot comparing Jev, Claude Haiku, and reactive rules for pick-and-place.
tryaksh/jev-pick-and-place-study · Python
10python - 1348jev-playgroundToy projects testing Jev (TypeSafe System One): snake autopilot, Todoist sorter, HN re-ranker, Dangerous Dave autopilot
nikhil1raghav/jev-playground · C
10c - 1350jev-playgroundA small Next.js app for experimenting with TypeSafe AI's Jev model (System One)
Little-Planet-Labs/jev-playground · TypeScript
11typescript - A WebGL demo where you play the card game Speed against a CPU whose brain is TypeSafe AI's Jev. The whole point of the app is to measure and show Jev's decision speed and decision accuracy in real time.
tubone24/jev-practice-speed · JavaScript
10card-gamejavascript - Review independente sobre a adequação do Jev a produtos e desenvolvimento multiagente
Teagar/jev-project-fit-review
10ai-routingdecision-model - automatically optimizing the instructions and decision criteria of TypeSafe Jev Choice from labeled data
j341nono/jev-prompt-optimization · Python
10python - 1358jev-relevancySearch evaluation with Jev as the judge. turbopuffer does the retrieval and hosted reranking, and DuckDB keeps every run and computes the metrics in SQL.
rickcrawford/jev-relevancy · Python
10bm25duckdbescihybrid-searchinformation-retrievalllm-as-a-judgendcg - 1359jev-report发明 RLHF 的人,这次做了个不会说话的模型:Jev 独立研究报告。52 页 PDF + 50 条中文实测复现包 + 143 条可回溯数据表
HackSing/jev-report · Python
10pythonai-researchchinesellm-evaluation - 1364jev-sandboxTest bench for TypeSafe's Jev
amr05008/jev-sandbox · TypeScript
10typescript - Benchmark of TypeSafe Jev (System One decision model) for systematic review title/abstract screening vs GLM/Sonnet on Cohen_2006 gold standard
Saeedabdf/jev-screening-benchmark · TypeScript
10typescript - 1368jev-showdownJev plays Pokemon Showdown: 40 battles, 940 decisions, reproducible results, and an annotated replay.
izzuddin8803/jev-showdown · Python
10ai-evaluationpokemon-showdownpython - 1369jev-simJev-compatible /v1/systemone server reading typed decisions from LLM logits, benchmarked against TypeSafe's Jev on the same items via JevBench
dashbi1/jev-sim · Python
10benchmarkcalibrationclassificationllmlogprobspython - Empirical zero-shot benchmark of TypeSafe Jev (v1.13.0) on low-resource Sinhala and Singlish NLP benchmarks (SOLD, SinhalaMMLU, NSINA)
Azri-Muhsin/jev-sinhala-review · Python
10benchmarkdecision-modelindic-nlpnlpsinhala-nlppython - Applying Jev to Snips Natural Language Understanding Benchmark
will-rice/jev-snips-benchmark · Python
11benchmarkevaluationfew-shotintent-classificationnatural-language-understandingnluslot-filling - 1390jev-testPre-registered benchmark: can a 2B local model (Gemma 4 E2B) answer web questions without making things up when a decision model (TypeSafe Jev) makes every call? SearXNG for search, MemPalace for verbatim memory, seven arms including open local judges. Spec and thresholds fixed before any run.
clduab11/jev-test · Python
10benchmarkgemmahallucinationmempalaceragsearxngsmall-language-models - 1392jev-testbedJev (TypeSafe System One) 테스트베드 — 클라우드 API와 로컬 셀프호스팅(jeff/GLiFormer) 양쪽 실행 예제 및 실측 결과
hulryung/jev-testbed · Python
10python - 1393jev-testbenchA lightweight, auditable Jev testbench built through Agentic Engineering, with frozen evidence, a preregistered null result, and a bilingual technical blog. 基于 Agentic Engineering 的轻量、可审计 Jev 实验台,包含冻结证据、预注册负结果、双语技术博客与认知负债反思。
Charlie-Wang-03/jev-testbench · Python
10agentic-engineeringai-agentsllmmodel-evaluationpreregistrationprobabilistic-aipython - 1416jevals-dataIndependent benchmark data for TypeSafe's Jev (System One model) vs LLMs: accuracy, calibration, cost. Boards + per-decision logs, CC-BY-4.0
Jevals/jevals-data
12benchmarkcalibrationdatasetleaderboardllm-evaluation - 1418jevbenchPreregistered, bias-corrected test of TypeSafe Jev's calibration under human disagreement (ChaosNLI, 100 labels per item)
GautamTalksDev/jevbench · Python
10calibrationchaosnlievaluationpreregistrationpython - 1439jevllmJev as an LLM (cz why not)
wisalkhanmv/jevllm · Python
10python - 1443jevoracleTyped questions over your own state, not a chat. A front end for TypeSafe's System One API.
tcsenpai/jevoracle · JavaScript
10buncalibrated-probabilitydecision-supportllm-alternativethreejs - 1447Jevs-GarageA garage full of tiny experiments for building critical systems with System One & Jev 🔧🧠⚡
JGalego/Jevs-Garage · Python
11ai-demosai-safetydecision-intelligencedeveloper-toolsexplainable-aihuman-in-the-loopincident-response - 1448JevScopeJevScope: TypeSafe Jev ecosystem repository.
KaushikKC/JevScope · TypeScript
10typescript - 1451jevtrafficsimTypeSafe AI's first model Jev takes on an entire city's traffic
skcache/jevtrafficsim · TypeScript
11typescript - 1465kyotsu-ai-benchAI benchmark on Japan's 2026 Common Test: Jev vs luna-none vs luna-low (static dashboard)
shibadogcap/kyotsu-ai-bench · HTML
10html - 1483moss-jevMOSS × Jev: a recorded-run 3D demo of the litter-picking rover choosing targets with TypeSafe's Jev decision model.
metrox-eth/moss-jev · JavaScript
12javascript - Typed-decision benchmark from PadFlow (land development SaaS): schemas, anonymized labeled rows, and a runner for confidence-calibrated models like TypeSafe Jev.
zsavage8/padflow-jev-evals · Python
10python - 1500paper-packageJev research manuscript, evidence, and reproducible paper package
Jev-Engineering/paper-package · Python
10 - 1523reflexreflex: TypeSafe Jev ecosystem repository.
conuko/reflex · TypeScript
10typescript - 1540stuntdoubleDrop-in /v1/systemone proxy that shadows Jev with local decision models (Kev, Laya) and reports whether you can swap
ReallyArtificial/stuntdouble · JavaScript
10ai-agentsevaluationgood-first-issuehelp-wantedjavascriptkevlaya - Rerunnable security experiments on System One decision models (TypeSafe Jev, Cloudflare Clef): state poisoning, prompt injection, truncation, and guard questions.
ankushchadha/system-one-security · TypeScript
10typescript - Recorded comparison of Jev, Terra, and Opus on 100 synthetic security-triage cases, five passes each, with a static inspectable dashboard
Robertzu43/system-one-security-triage · TypeScript
10ai-securitybenchmarkllm-evaluationsecuritytypescript - typesafe-jev-traffic-demo: TypeSafe Jev ecosystem repository.
trycatchkamal/typesafe-jev-traffic-demo · Python
10python - Next.js UI showing off TypeSafe's System One model (Jev) — parallel Noul judgments and a Choice-based citation checker, deployable to Vercel
Ashadeepa/typesafe-showcase · TypeScript
10typescript - Ultimate Agents - Space Invaders - A 27B parameter LLM loses to pressing FIRE and code beats everyone
13rac1/ufa-2026-09-spaceinvaders · Python
10python - 1629AnyLM2JevReproducing the Jev decision interface
our0boros/AnyLM2Jev · Python
00calibrationllmpython - Triage real wind turbine SCADA events with Jev typed questions, with a second look from production data for uncertain events
tameernoor/aotn-jev-turbine-triage · Python
00python
151–200 of 389