Defines MIIT, constructs MIIT-dataset via automatic pipeline across seven risk categories, and trains MiShield-8B which outperforms commercial moderation services using progressively distilled reasoning supervision.
Baseline reference
Mmiu: Multimodal multi-image understanding for evaluating large vision-language models
Baseline reference. 50% of citing Pith papers use this work as a benchmark or comparison.
10
Pith papers citing it
Baseline
50% of classified citations
citation-role summary
background 3
dataset 3
citation-polarity summary
representative citing papers
CGC improves fine-grained multi-image understanding in MLLMs by constructing contrastive training instances from existing single-image annotations and adding a rule-based spatial reward, achieving SOTA on MIG-Bench and VLM2-Bench with transfer gains to other multimodal tasks.
Reformulating 53 visual reasoning tasks in polar coordinates causes frontier MLLMs to drop from 70-83% to 31-39% accuracy while preserving logical equivalence, revealing a Cartesian shortcut in current benchmarks.
OMIBench benchmark reveals that current LVLMs achieve at most 50% on Olympiad problems requiring reasoning across multiple images.
PulseFocus improves multi-image reasoning in VLMs by interleaving planning and attention-gated focus blocks during chain-of-thought, achieving gains on BLINK and MuirBench.
InternVL3.5 advances open-source multimodal models with Cascade RL for +16% reasoning gains and ViR for 4x inference speedup, with the 241B model reaching SOTA among open-source MLLMs on multimodal, reasoning, and agentic tasks.
InternVL3-78B sets a new open-source SOTA of 72.2 on MMMU via native joint multimodal pre-training, V2PE, MPO, and test-time scaling while remaining competitive with proprietary models.
InternVL 2.5 is the first open-source MLLM to surpass 70% on the MMMU benchmark via model, data, and test-time scaling, with a 3.7-point gain from chain-of-thought reasoning.
InternVideo3 introduces Multimodal Contextual Reasoning and M^2LA attention to enable closed-loop evidence accumulation in long-video understanding and agentic tool use, reporting strong benchmark results.
A pipeline of chain-of-thought data synthesis, LoRA-based supervised fine-tuning, rejection sampling, and rule-based reinforcement learning raises multi-image grounding accuracy by 9.04% on MIG-Bench and 4.41% on average across seven other benchmarks.
citing papers explorer
-
InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models
InternVL3-78B sets a new open-source SOTA of 72.2 on MMMU via native joint multimodal pre-training, V2PE, MPO, and test-time scaling while remaining competitive with proprietary models.