Open decision models and reproductions with decision models

Open Apache-2.0 System One decision models (4B, 9B, 27B on Qwen) with trainer, data builders, evals and every run record.

PostHog's 9B Jev-like model that reasons before deciding, trained with SFT and CISPO, with a diffusion drafter for speed.

Minimal training project for a Jev-style decision model: Qwen3-0.6B with a pointer head that outputs probabilities over options.

Open multimodal decision model fine-tuned from Gemma 4 12B that returns option probabilities for text, image, audio and video questions.

Research router where a Qwen3-1.7B belief model scores intent candidates and a small policy decides to commit, retrieve more or hand off.

Preprint survey of Jev-style typed decision models: their research roots, adaptation costs and what current evidence supports.

Open infrastructure for training, evaluating and deploying System 1 decision models, with calibrated LoRA checkpoints on a 27B backbone.

Open stack to synthesise decision data, train and calibrate a Qwen3.5-0.8B scorer, evaluate on Mind2Web and serve Jev-style.

Training library for a small one-pass model that scores a changing list of text options and returns a probability per option.

Open Jev-style 0.6B decision model plugged into BlitzRank and Reranker-Guided Search, scoring k candidates per match in one pass.

Clojure System One engine on Jolt that routes typed questions to Laya encoders, a llama.cpp 'thinker', Clef or TinyDecide.

Reproduction of Jev's decision layer on any Qwen model: serves /v1/systemone by scoring option tokens, with no training or generation.

Offline System One server that answers Jev-format questions with small open models, wire-compatible with TypeSafe's official SDK.

Image-plus-text decision model on Qwen3-VL-2B with LoRA and an answer head that returns one probability per option.

Open MiniCPM5-2B decision model serving /v1/systemone: typed choice, noul and score probabilities in one forward pass.

Jev-style replica: frozen Qwen3-1.7B plus a ~0.8M-parameter trained head returns calibrated Choice, Boolean and Score distributions.

Browser lab and TypeSafe-shaped API for asking typed questions of a Nemotron diffusion model and inspecting the distributions.

nev-lite: 11.8M-parameter static-embedding decision head that scores fixed option sets with no transformer pass at inference.

OmniJev: omni-modal 4B decision model answering typed questions about images, video, screens and robot scenes in one pass.

Open 150M encoder that answers noul, choice and score questions in one pass, trainable on a free Colab T4 in about 30 minutes.

Two frozen open models answer yes/no per exam option; their agreement drives an auto-release gate and a conformal candidate set.

spark-s1: local Jev-inspired decision models on Qwen3.5-4B for NVIDIA DGX Spark, trained on code- and solver-labelled decision rows.

Drop-in faster backend for Open-Jev-27B using custom CUDA kernels, a prefix tree and CUDA graphs: about 6x lower JevBench latency.

Verdict: 151M ModernBERT/GLiClass non-autoregressive decision models with calibration heads, reporting wins over Jev and Laya.

Research toolkit that encodes an image once and answers several typed questions from a shared distribution, plus OpenJev-0.6B text models.

Phocinae Largha: 144.3M bilingual English/Chinese decision model on mmBERT-small with calibrated Choice, Score and Noul in one pass.

PlayJev: Qwen3.5-0.8B fine-tuned to play ten browser games from raw pixels, one forward pass per move with a probability per option.

Open decision models fine-tuned from Qwen at 0.8B to 9B that return a choice and per-option probabilities for Choice, Noul and Score.

Prototype Jev-style decision model on Qwen3.5-0.8B using prefix-fork execution to answer typed questions in one pass without generation.

Reproducible MLX and Colab recipes that turn Qwen decoders into Jev-like typed decision models, from 0.5B on a laptop to a noncausal 27B.