Skip to content
Trending storyDeveloping

ModelScope releases open-weight JEV-27B-VL vision model, ranked first on Decision Index Vision

1 article1 sourcesince Oct 10Last article 1h ago ·

Overview

AISummary of 1 article

ModelScope releases JEV-27B-VL, an open-weight vision-language model under Apache 2.0 that ranks first among 20 models on the independent Decision Index Vision board.

ModelScope reports 77.8% accuracy across 13,695 examples on the board.

According to the source, the model's confidence scores closely track real accuracy, with an ECE of 0.019, and it accepts image-and-text prompts up to 256K tokens. The source also says its decision adapter, trained only on text, exceeds the 397B Qwen3.5 reference by 6.4 points on the board's full score in zero-shot vision.

Written by AI from the articles below · updated Oct 10, 5:55 AM ET

Check the sources:

Latest developmentOct 10, 5:30 AM ET
JEV-27B-VL ranks first on Decision Index Vision with calibrated visual decisions

Article timeline

The articles in this story. Times are ET.

Oct 10
  1. ModelScopeOfficial
    JEV-27B-VL ranks first on Decision Index Vision with calibrated visual decisions

    AIAutoTrust Labs releases JEV-27B-VL, which ranks first among 20 models on the independent Decision Index Vision board with 77.8% accuracy across 13,695 examples. Its confidence scores closely track real accuracy, with an ECE of 0.019, and it supports image-and-text prompts up to 256K tokens. The model is released with open weights under Apache 2.0, and its decision adapter, trained only on text, exceeds the 397B Qwen3.5 reference by 6.4 points on the board's full score in zero-shot vision.

    Video from @ModelScope2022's post

Heat trend

Not enough continuous observations to show a trend yet.