Nnican2018Benchmarks & research

shade-arena-jev-monitor

Evaluating TypeSafe's Jev as a fast monitor and action gate for agent sabotage in SHADE-Arena, compared with Gemini 2.5 Flash/Pro.

Explore repository

How it uses Jev

SHADE-Arena + Jev monitor evaluation

Getting started

On Windows, run the scripts inside the activated environment (or with conda run ). Numpy/MKL cannot find its DLLs when the environment's python.exe is called directly.

Source: community catalog. Details and metrics reflect the published community snapshot and are not a performance endorsement. Documentation details extracted from GitHub · the README.

Keep exploring

Similar to shade-arena-jev-monitor

Browse all
jev-column-race project preview

jev-column-race

Benchmarks & research
Jev vs Gemini 3.8 Flash: labelling 1,000 app reviews, 4.1× faster and 7× cheaper
GMade bygoodrahstar23
jev-agent-failure-benchmark project preview

jev-agent-failure-benchmark

Benchmarks & research
Benchmarking Jev (Typesafe.ai) against a strong LLM on the Who&When Pro agent-failure-attribution benchmark (text subset).
TMade byTokenTrim2
jev-anotacao-sentencas project preview

jev-anotacao-sentencas

Benchmarks & research
Jev (TypeSafe) vs. Gemini 3.8 Flash vs. GPT-5.6 Luna na anotação estruturada de sentenças do TJSP: qualidade, tempo e custo
LMade bylab-dados0
typesafe-arena project preview

typesafe-arena

Demos & playgrounds
A playground for TypeSafeAI's Jev Model
DMade byDeepBlueDynamics2
tempo-jev-demo project preview

tempo-jev-demo

Demos & playgrounds
A natural-language task workspace comparing performance across AI models (TypeSafe's Jev, GPT-5.6 Luna, and Gemini 3.8 Flash)
MMade bymychaelangelo2
jev-arena-nanojev project preview

jev-arena-nanojev

Games & simulations
完全本地的 NanoJev 网格决策游戏实验场，支持中文 Pygame、多关卡与 GTX 1660S 训练
LMade byliao963124
agi-jev-containment project preview

agi-jev-containment

Games & simulations
AGI JEV Detection — local AI agent monitor: chain-level malicious-agent detection (TypeSafe Jev + Sentinel), escalate-only L1–L5 containment, Neo4j forensics, AngryRobot dashboard. HackSpain 2026.
CMade bycarlosedm104
Augustus project preview

Augustus

Agent tooling
Independent agent skill for finding, building, evaluating, and improving decision-model systems with composition rules, evaluation harnesses, and bounded prompt/program optimization; TypeSafe Jev is the default hosted exemplar.
2Made by246019
muse-jev-playbook project preview

muse-jev-playbook

Agent tooling
Jev decision layer for Muse: a fast, cheap TypeSafe AI gate before expensive agent work — confidence policy, recipes, reference router, honest measurement.
BMade byBodila5116