







Evaluates typed decisions (choice, score, noul) over 100+ languages in a single forward pass with calibrated probabilities. Outperforms TypeSafe Jev.
GitHub - AbdelStark/jev-benchmarks at 0d610cc53e79bcbec691312b0c4adb4a0e371642
Probability-aware evaluation for typed decision models: calibration, selective risk, latency, and reproducible benchmarks. - AbdelStark/jev-benchmarks

Mitigating Cross-Lingual Cultural Inconsistencies in LLMs via...
Despite their impressive capabilities, multilingual large language models (MLLMs) frequently exhibit inconsistent behaviour when the prompt's language changes. While such adaptation is generally...

chad/whichlang
What programming language do LLMs default to when you don't tell them? A small benchmark.
JevBench — AI decision model benchmark & leaderboard
Compare AI decision models on accuracy, probability calibration, latency and cost. JevBench by Benchmark Heaven ranks open-weights decision models by the JevBench Capability Score, with methodology and reproducible aggregate results.
Targeted Multilingual Adaptation for Low-resource Language Families
The "massively-multilingual" training of multilingual models is known to limit their utility in any one language, and they perform particularly poorly on low-resource languages. However, there is evid
System One: fast judgments and deliberate checks
Explore System One thinking, its limits, and how fast judgments can be used in software. Interactive examples, with Jev as a case study.
Jev introduces a new shape of LLM—System One, aka Decision Models
Last week TypeSafe AI unveiled Jev, their first example of a new category of model that they are calling “System One models” (I’m with Maggie Appleton, I think “decision models” …
How Large Language Models Actually Work
Computation and its Connotations
A Review of Language Machines by Leif Weatherby

ChatGPT Translate | Fast, Natural, 40+ Languages
ChatGPT translates across 40+ languages with accuracy, tone, and cultural nuance. Translate text, voice, or photos for everyday use, travel, school, and work — and learn grammar or phrasing as you go.

Please
Please is a cross-language build system with an emphasis on high performance, portability, extensibility and correctness.

On-Device LLM Throughput Calculator - a Hugging Face Space by FL33TW00D-HF
This tool estimates and visualizes the throughput of Large Language Models on devices with memory bandwidth constraints. Users input device and model configurations, and the tool generates a plot s...
inanna-malick/jev-dsl
Agent-first Haskell DSL for TypeSafe's Jev judgment model: typed packets, inferred types, answers under the same labels

Perplexity Decider v1.3: Training, eval, and use cases
How we train and evaluate Decider v1.3 and Decider Flash, with locate extraction, real-world results, and costs.
