User repository on GitHub
- Forks
- 2
- Last pushed
- Updated Sep 2026
Repo stats from the GitHub API, cached Oct 2026.
CommunityJev Models & Reimplementations
Visual-Jev
A Jev-style typed-decision baseline built on the frozen Qwen3.5-4B checkpoint that accepts a local image path directly, letting the vision-language model score declared options without generating text. Ships scoring modes, a WebGPU browser demo and benchmark results measured on an RTX 3090.
- Stars
- 9
- Language
- Python
- License
- MIT
- Published by
- Community
- Last verified
- Oct 2026
About this project
Semantic ifs from open models, on a 3090 at home. Independent project; not affiliated with Jev or TypeSafe. Wow! No waitlist. Run it in your browser today. Same frozen 4B model · same state · same 21 questions · measured separately, aligned at t=0 in the replay. Most agent decisions are small: route this, retry that, does the evidence support X? A chat model can answer them, but it spends time generating text that software immediately parses back into an if statement.
Quoted from the project's README. Read the full README.
- Languages
- Python
- Capabilities
- Open sourceRuns locally
Highlights
- Optional `Image` path runs Qwen3.5-4B as a vision-language model; state, question and options stay the same and no tokens are generated.
- Direct logit readout answered 21 criteria in 1.023 s versus 5.332 s for an autoregressive JSON array on one RTX 3090.
- Shared-state scoring reached 20.03 decisions/s over a 37x21 fixture, versus 2.33 decisions/s for fresh direct scoring.
- Browser WebGPU demo runs a 3.01 GB quantized Qwen3.5-4B build with 0.813 authored balanced accuracy.
- Independent project, formerly OpenJev, that the README says does not reproduce Jev's undisclosed model or training.
Watch out
MIT-licensed research baseline, but model weights and third-party sources are not included in the repository, and quality and speed numbers are the authors' own measurements on their hardware.
Questions about this entry
Short answers drawn from the linked project or page.
Add a nonempty local Image path to a decision row. The processor reads the file directly and the model still returns option logits in one forward pass, with no answer sentence, JSON repair or decoding loop. Text-only rows are unchanged.
No. The README states it is an independent research project that reproduces only the typed-decision interface pattern with open Qwen3.5 models, and it does not reproduce Jev's model or training.
Python 3.10+, CUDA and a GPU that can hold a 4B BF16 model, installed with pip install -e '.[test]'. On Apple Silicon, install .[test,mlx] and add --backend mlx to the scorer command for direct, serial and shared scoring.
direct, serial prefix reuse, shared-state parallel and a native reranker mode. If every row has the same exact state, --mode shared prefills it once and evaluates the criteria in parallel. MLX and --mode reranker reject Image rows.
Typed option scores, timing, the exact model revision and a prompt hash. The scorer refuses an existing output path, so pick a new --output file on every run. Returned probabilities are conditional on the options you supplied, so calibrate them on your own workload.
On the 102-row TypeSafe subset that could be aligned from public artifacts, the browser Qwen3.5-4B build agreed with the modal answer 0.845 of the time against Jev's published 0.883. The Jev number is read from TypeSafe's published records, not a live endpoint.
Yes. A WebGPU demo runs a quantized 3.01 GB Qwen3.5-4B build with 0.813 authored balanced accuracy and no waitlist. The same browser ladder also benchmarks Qwen3-0.6B at 0.440 and MiniCPM5-2B at 0.686.
Reactions & coverage
Posts, threads, and videos about this entry from around the web.
LinkedIn: What multimodal decision models actually change, and what they do not.
LinkedInJevDirectory.org
Hacker News: Decision Index 0.2.1 scores open reproductions of Jev on a frozen suite.
jackbravo4 pointsDecision Index 0.2.1 – 70 open reproductions of TypeSafe Jev's Decision ModelRead the thread on Hacker NewsHacker News: Pointer to a DiffusionGemma plus vLLM tutorial for a Jev-shaped classifier.
1 pointHow to build a Jev-style classifier with DiffusionGemma and vLLMRead the thread on Hacker NewsHacker News: A one-token logprob wrapper that also classifies webcam frames.
allanrbo151 pointsA single function Jev-like wrapper for LLMs, including vision modelsRead the thread on Hacker NewsMore Jev resources like this
1.3kGitHub stars
A local yes/no decision model that speaks Jev's wire format on a laptop CPU, refusing Choice and Score until those question types ship.
Jev Models & Reimplementations#community#python#open-models
Community
feder-cr/jevC++
feder-cr/jevC++The vision sibling: the same text-trained System 1 head on Qwen3.8-27B's image encoder, taking images zero-shot, ranked first of 20 vision decision models on the community Jev Decision Index 0.3.
Jev Models & Reimplementations#community#open-models#local
Community
A 400M-parameter Jev-compatible decision model with a browser build: 0.796 on typed decisions against Laya's 0.766, 17 ms per question on an L40S, and an ONNX file that runs in a tab.
Jev Models & Reimplementations#community#open-models#local
Community
4.8kGitHub stars
An open-model semantic-if engine positioned as a local Jev-style decision stack, separate from the original SemIf repo but sharing the same author.
Jev Models & Reimplementations#community#python#open-models
Community
TheoLeeCJ/SemIf-OpenJevPython
TheoLeeCJ/SemIf-OpenJevPython877GitHub stars
A local server that reads next-token probabilities from a fine-tuned Spark model and returns Choice, Score, and Noul answers with zero generated tokens.
Jev Models & Reimplementations#community#python#open-models
Community
Rizzo-AI-Academy/rizzo-flowPython
Rizzo-AI-Academy/rizzo-flowPython738GitHub stars
A multimodal decision model that scores text, image, and video candidates with a shared head on Qwen3.5, plus SFT and experimental RLCD training code.
Jev Models & Reimplementations#community#python#open-models
Community
Liuziyu77/ValenPython
Liuziyu77/ValenPythonFrom the Jev guides
Original write-ups on what this entry is for.
Jev takes text only; Clef-omni adds image, audio, and video input and Liquid d1 adds images, each keeping the same typed-answer shape. Where that matters, where it does not, and the costs the vendors underplay.
Jev Models & Reimplementations#multimodal#open-models#decision-model
Read article
One model in September, a dozen by October. Clef, Liquid d1, Perplexity, and OpenAI all ship the same typed-answer shape now. How to read the rival benchmarks and pick on your own task.
Jev Models & Reimplementations#decision-model#benchmarks#pricing
Read article
Hundreds of repositories carry the system-one tag, but the ecosystem is a dozen ideas in different sizes. A tour of the families, the licenses, and what a download does not include.
Jev Models & Reimplementations#open-models#hugging-face#local
Read article
From the Jev community
Posts from builders shipping with Jev right now.