Browser Use + Jev
Gregor Zunic's flight-search demo with a dynamic DOM action space.
A FIELD GUIDE TO JEV / VOL. 01
The best of Jev. Real projects, practical guides and ideas from across the internet.
Gregor Zunic's flight-search demo with a dynamic DOM action space.
01 / A small decision engine. A whole new way to build.
114 resources / The collection
Reset filtersAI-assisted summaries and translations. Check original sources for context and performance claims.
A chatbot from typed Jev decisions: hierarchical speculative decoding over System One probabilities.
Jev monitors 25 WhatsApp customer groups in real time, alerting an LLM only when urgent issues arise.
Jev (TypeSafe) exploratory thread: claim audit, live demos, and runnable code.
An observable raw-character chat experiment powered entirely by TypeSafe Jev Choice.
A research project on GitHub uses Jev with Qwen3 models on an NVIDIA DGX Spark system.
Probability-aware evaluation for typed decision models: calibration, selective risk, latency, and reproducible benchmarks.
A stronger one-pass scorer over a variable list of text options: hashed n-gram encoder, rival-aware attention, gated head, temperature scaling, benchmarked against jevlike.
Ongoing Japanese research deck on Jev and System One models, maintained as Markdown slides.
Feedback on your paper in seconds.
The open-source System One decision model. Sub-15ms, non-autoregressive, local drop-in alternative to TypeSafe Jev.
AI benchmark on Japan's 2026 Common Test: Jev vs luna-none vs luna-low (static dashboard).
Zero-shot spam filtering with TypeSafe Jev Noul questions, compared with TF-IDF baselines.
Replicating Jev with a local LLM.
One-pass typed decisions with calibrated probabilities (System One style model), fine-tuned from Qwen3.5-2B.
Open replica of TypeSafe's Jev: typed calibrated decisions in one forward pass, on Gemma 4 E2B / Gemma 3 270M (Modal).
Local bilingual probability decisions from context, questions, and candidate answers. Independent research preview inspired by TypeSafe Jev.
Jev integrates into a recommendation engine, providing 70+ personalized suggestions per user at 0.0019$ per user.
Openvons (open-Jev): 有限選択肢に確率で答える判断層 — テキスト / 画像 / 日本語音声コマンド.
Vercel Labs terminal CLI that can run Jev as the evaluation model for its evaluate command.
A reproduction of Jev that turns any Qwen model into a fast decision model, serving the same /v1/systemone schema (Choice, Score, Noul) with no training and no generated answer text.
A real-time tool to analyze and categorize tweets for viral potential.
TypeSafe Jev demonstration for new analyzation — experimenting with Jev for fast analysis of news and tickers.
A PostgreSQL extension enabling natural language queries across databases.
High-throughput synthetic & pretraining dataset sifter powered by TypeSafe AI Jev (api.typesafe.ai). Stream, filter, and score Parquet & JSONL datasets at 1,500+ rows/sec using System One typed decisions (Choice, Score, Noul).