CVPR25-MLLM-Paper-List

CVPR 2025 Multimodal Large Language Models Paper List

📖 Table of Contents

• 📖 Table of Contents
  • Image LLMs
  • Video LLMs
  • Unified LLMs
  • Other Modalities
  • Preference Optimization
  • Jailbreak
  • Benchmarks
  • Retrieval

Image LLMs

• LLaVA-Critic: Learning to Evaluate Multimodal Models Paper Page
• Img-Diff: Contrastive Data Synthesis for Multimodal Large Language Models Paper Code
• FlashSloth: Lightning Multimodal Large Language Models via Embedded Visual Compression Paper Code
• BlueLM-V-3B: Algorithm and System Co-Design for Multimodal Large Language Models on Mobile Devices Paper Code
• Insight-V: Exploring Long-Chain Visual Reasoning with Multimodal Large Language Models Paper Code
• Critic-V: VLM Critics Help Catch VLM Errors in Multimodal Reasoning Paper
• Mono-InternVL: Pushing the Boundaries of Monolithic Multimodal Large Language Models with Endogenous Visual Pre-training Paper Code
• DivPrune: Diversity-based Visual Token Pruning for Large Multimodal Models Paper Code
• ODE: Open-Set Evaluation of Hallucinations in Multimodal Large Language Models Paper
• Augmenting Multimodal LLMs with Self-Reflective Tokens for Knowledge-based Visual Question Answering Paper Code
• AGLA: Mitigating Object Hallucinations in Large Vision-Language Models with Assembly of Global and Local Attention Paper Code
• ICT: Image-Object Cross-Level Trusted Intervention for Mitigating Object Hallucination in Large Vision-Language Models Paper
• Can Large Vision-Language Models Correct Grounding Errors By Themselves? Paper
• Molmo and PixMo: Open Weights and Open Data for State-of-the-Art Vision-Language Models Paper Page
• Nullu: Mitigating Object Hallucinations in Large Vision-Language Models via HalluSpace Projection Paper Code
• HiRes-LLaVA: Restoring Fragmentation Input in High-Resolution Large Vision-Language Models Paper
• Devils in Middle Layers of Large Vision-Language Models: Interpreting, Detecting and Mitigating Object Hallucinations via Attention Lens Paper
• HoVLE: Unleashing the Power of Monolithic Vision-Language Models with Holistic Vision-Language Embedding Paper
• VoCo-LLaMA: Towards Vision Compression with Large Language Models Paper Code
• Perception Tokens Enhance Visual Reasoning in Multimodal Language Models Paper Page
• Florence-VL: Enhancing Vision-Language Models with Generative Vision Encoder and Depth-Breadth Fusion Paper Page
• ATP-LLaVA: Adaptive Token Pruning for Large Vision Language Models Paper Page
• Accelerating Multimodel Large Language Models by Searching Optimal Vision Token Reduction Paper
• VisionZip: Longer is Better but Not Necessary in Vision Language Models Paper Code

Video LLMs

• Apollo: An Exploration of Video Understanding in Large Multi-Modal Models Paper Page
• VideoRefer Suite: Advancing Spatial-Temporal Object Understanding with Video LLM Paper Code
• Seq2Time: Sequential Knowledge Transfer for Video LLM Temporal Grounding Paper
• On the Consistency of Video Large Language Models in Temporal Comprehension Paper Code
• DyCoke: Dynamic Compression of Tokens for Fast Video Large Language Models Paper Code
• PAVE: Patching and Adapting Video Large Language Models Paper Code
• DynFocus: Dynamic Cooperative Network Empowers LLMs with Video Understanding Paper Code
• M-LLM Based Video Frame Selection for Efficient Video Understanding Paper
• Adaptive Keyframe Sampling for Long Video Understanding Paper Code
• VISTA: Enhancing Long-Duration and High-Resolution Video Understanding by Video Spatiotemporal Augmentation Paper Code
• LLaVA-ST: A Multimodal Large Language Model for Fine-Grained Spatial-Temporal Understanding Paper Code
• Unlocking Video-LLM via Agent-of-Thoughts Distillation Paper Page
• STEP: Enhancing Video-LLMs’ Compositional Reasoning by Spatio-Temporal Graph-guided Self-Training Paper
• PVC: Progressive Visual Token Compression for Unified Image and Video Processing in Large Vision-Language Models Paper Code
• NVILA: Efficient Frontier Visual Language Models Paper Code
• Video-Panda: Parameter-efficient Alignment for Encoder-free Video-Language Models Paper Code
• HierarQ: Task-Aware Hierarchical Q-Former for Enhanced Video Understanding Paper Page

Unified LLMs

• Janus: Decoupling Visual Encoding for Unified Multimodal Understanding and Generation Paper Code
• JanusFlow: Harmonizing Autoregression and Rectified Flow for Unified Multimodal Understanding and Generation Paper Code
• MMAR: Towards Lossless Multi-Modal Auto-Regressive Probabilistic Modeling Paper Code
• CoMM: A Coherent Interleaved Image-Text Dataset for Multimodal Understanding and Generation Paper Code
• WeGen: A Unified Model for Interactive Multimodal Generation as We Chat Paper Code
• SynerGen-VL: Towards Synergistic Image Understanding and Generation with Vision Experts and Token Folding Paper
• TokenFlow: Unified Image Tokenizer for Multimodal Understanding and Generation Paper Code

Other Modalities

• Thinking in Space: How Multimodal Large Language Models See, Remember and Recall Spaces Paper Code
• EventGPT: Event Stream Understanding with Multimodal Large Language Models Paper Code

Preference Optimization

• Task Preference Optimization: Improving Multimodal Large Language Models Performance with Vision Task Alignment Paper Code
• SymDPO: Boosting In-Context Learning of Large Multimodal Models with Symbol Demonstration Direct Preference Optimization Paper Code
• Continual SFT Matches Multimodal RLHF with Negative Supervision Paper
• Mitigating Hallucinations in Large Vision-Language Models via DPO: On-Policy Data Hold the Key Paper Code

Jailbreak

• Immune: Improving Safety Against Jailbreaks in Multi-modal LLMs via Inference-Time Alignment Paper Code
• Distraction is All You Need for Multimodal Large Language Model Jailbreaking Paper
• Playing the Fool: Jailbreaking LLMs and Multimodal LLMs with Out-of-Distribution Strategy Paper
• Exploring Visual Vulnerabilities via Multi-Loss Adversarial Search for Jailbreaking Vision-Language Models Paper
• Steering Away from Harm: An Adaptive Approach to Defending Vision Language Model Against Jailbreaks Paper Code

Retrieval

• GME: Improving Universal Multimodal Retrieval by Multimodal LLMs Paper Code
• Retrieval-Augmented Personalization for Multimodal Large Language Models Paper Code

Benchmarks

• Video-MME: The First-Ever Comprehensive Evaluation Benchmark of Multi-modal LLMs in Video Analysis Paper Code
• MLVU: Benchmarking Multi-task Long Video Understanding Paper Code
• MMVU: Measuring Expert-Level Multi-Discipline Video Understanding Paper Code
• MV-MATH: Evaluating Multimodal Math Reasoning in Multi-Visual Contexts Paper Page
• VideoAutoArena: An Automated Arena for Evaluating Large Multimodal Models in Video Analysis through User Simulation Paper Code
• OVBench: How Far is Your Video-LLMs from Real-World Online Video Understanding? Paper Code
• ECBench: Can Multi-modal Foundation Models Understand the Egocentric World? A Holistic Embodied Cognition Benchmark Paper Code
• Localizing Events in Videos with Multimodal Queries Paper Code
• ICQ: Is `Right' Right? Enhancing Object Orientation Understanding in Multimodal Language Models through Egocentric Instruction Tuning Paper Code
• VidHalluc: Evaluating Temporal Hallucinations in Multimodal Large Language Models for Video Understanding Paper Page
• VidComposition: Can MLLMs Analyze Compositions in Compiled Video? Paper Code
• Unveiling the Ignorance of MLLMs: Seeing Clearly, Answering Incorrectly Paper
• VL-RewardBench: A Challenging Benchmark for Vision-Language Generative Reward Models Paper Page
• Benchmarking Large Vision-Language Models via Directed Scene Graph for Comprehensive Image Captioning Paper Code
• OpenING: A Comprehensive Benchmark for Judging Open-ended Interleaved Image-Text Generation Paper Code
• Automated Generation of Challenging Multiple-Choice Questions for Vision Language Model Evaluation Paper Page
• VideoChat-Online: Towards Online Spatial-Temporal Video Understanding via Large Video Language Models Paper Page
• ODE: Open-Set Evaluation of Hallucinations in Multimodal Large Language Models Paper
• VideoEspresso: A Large-Scale Chain-of-Thought Dataset for Fine-Grained Video Reasoning via Core Frame Selection Paper Code