mlsanityData to 7 Oct 2026

MJ1: Multimodal Judgment via Grounded Verification

arXiv 2603.07990 ↗First posted 9 Mar 2026cs.LGcited by 2

Not on the map. Why?

Abstract (arXiv)

Multimodal judges struggle to ground decisions in visual evidence. We present MJ1, a multimodal judge trained with reinforcement learning that enforces visual grounding through a structured grounded verification chain (observations rightarrow claims rightarrow verification rightarrow evaluation rightarrow scoring) and a counterfactual consistency reward that penalizes position bias. Even without training, our mechanism improves base-model accuracy on MMRB2 by +3.8 points on Image Editing and +1.7 on Multimodal Reasoning. After training, MJ1, with only 3B active parameters, achieves 77.0% accuracy on MMRB2 and surpasses orders-of-magnitude larger models like Gemini-3-Pro. These results show that grounded verification and consistency-based training substantially improve multimodal judgment without increasing model scale.

Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 18 of its 26 references that are in the map. First posted on arXiv 9 Mar 2026.

Builds on

18 of 26 references.

References · 18

  1. 2022Training language models to follow instructions with human feedbackcited by 15,712
  2. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 8,851
  3. 2023Judging LLM-as-a-Judge with MT-Bench and Chatbot Arenacited by 8,393
  4. 2025Qwen3 Technical Reportcited by 8,195
  5. 2025Qwen2.5-VL Technical Reportcited by 5,418
  6. 2024GPT-4o System Cardcited by 4,467
  7. 2020Shortcut Learning in Deep Neural Networkscited by 1,928
  8. 2025Gemma 3 Technical Reportcited by 1,686
  9. 2023Measuring Faithfulness in Chain-of-Thought Reasoningcited by 668
  10. 2024An Image is Worth 1/2 Tokens After Layer 2: Plug-and-Play Inference Acceleration for Large Vision-Language Modelscited by 629
  11. 2023OPERA: Alleviating Hallucination in Multi-Modal Large Language Models via Over-Trust Penalty and Retrospection-Allocationcited by 468
  12. 2024SparseVLM: Visual Token Sparsification for Efficient Vision-Language Model Inferencecited by 378
6 more
  1. 2025LLaVA-Mini: Efficient Image and Video Large Multimodal Models with One Vision Tokencited by 164
  2. 2025JudgeLRM: Large Reasoning Models as a Judgecited by 76
  3. 2025Learning to Plan & Reason for Evaluation with Thinking-LLM-as-a-Judgecited by 75
  4. 2025J1: Incentivizing Thinking in LLM-as-a-Judge via Reinforcement Learningcited by 73
  5. 2025Multimodal RewardBench: Holistic Evaluation of Reward Models for Vision Language Modelscited by 40
  6. 2024Do Vision-Language Models Really Understand Visual Language?cited by 29

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 2.

2 papers.

  1. 2026A Cross-Model VLM-Judge Protocol for Single-Image 3D Mesh Quality (and Why Cheap Proxies Fall Short)cited by 2
  2. 2026A Low Grounding Score Is Not an Ungrounded Judge: Identifying the Perceptibility Confound in Multimodal Oversightcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026DeltaRubric: Generative Multimodal Reward Modeling via Joint Planning and Verification
  2. 2026Faithful-MR1: Faithful Multimodal Reasoning via Anchoring and Reinforcing Visual Attention
  3. 2026See It, Say It, Sorted: An Iterative Training-Free Framework for Visually-Grounded Multimodal Reasoning in LVLMs
  4. 2025Multi-Crit: Benchmarking Multimodal Judges on Pluralistic Criteria-Following
  5. 2026Mitigating Perceptual Judgment Bias in Multimodal LLM-as-a-Judge via Perceptual Perturbation and Reward Modeling
  6. 2026Reinforcing Multimodal Reasoning via Token-Level Perception-Grounded Advantage Estimation
  7. 2025MMRPT: MultiModal Reinforcement Pre-Training via Masked Vision-Dependent Reasoning
  8. 2026Judge Model for Large-scale Multimodality Benchmarks