Tags: Multimodal Large Language Model - Paper Library

UniWorld-V1: High-Resolution Semantic Encoders for Unified Visual Understanding and Generation

Published:6/3/2025

Unified Visual Understanding and Generation ModelHigh-Resolution Semantic EncodersContrastive Semantic Encoding-Based Generative FrameworkImage Understanding and GenerationMultimodal Large Language Model

UniWorldV1 is an innovative generative framework that integrates highresolution semantic encoders for visual understanding and generation, achieving impressive performance across tasks using only 2.7M training data.

ThinkAct: Vision-Language-Action Reasoning via Reinforced Visual Latent Planning

Published:7/23/2025

Vision-Language-Action ReasoningReinforced Visual Latent PlanningMultimodal Large Language ModelLong-Horizon PlanningRobotic Action Execution

ThinkAct proposes a dualsystem framework that connects highlevel reasoning and lowlevel action execution through reinforced visual latent planning, enabling fewshot adaptation, longhorizon planning, and selfcorrection in complex environments.

A Multi-modal Large Language Model with Graph-of-Thought for Effective Recommendation

Published:1/1/2025

Multimodal Large Language ModelGraph-of-Thought Prompting TechniquePersonalized Recommendation SystemMultimodal Recommendation TasksUser-Item Interaction Graphs

The GollaRec model integrates a Multimodal Large Language Model and GraphofThought to enhance useritem interaction for effective recommendations, combining visual and textual data. It utilizes textgraph alignment and tuning, outperforming 12 existing models in multimodal tas

Qwen3-Omni Technical Report

Published:9/22/2025

Multimodal Large Language ModelQwen3-Omni ArchitectureAudio Task Performance OptimizationGeneral-Purpose Audio CaptioningMultilingual Speech Understanding and Generation

Qwen3Omni is a single multimodal model achieving stateoftheart performance across text, image, audio, and video, particularly excelling in audio tasks. It uses a mixtureofexperts architecture, supports multilingual audio understanding and generation, and reduces latency wit

ADriver-I: A General World Model for Autonomous Driving

Published:11/23/2023

Driving World ModelsMultimodal Large Language ModelVision-Language-Action ModelDiffusion ModelsnuScenes Dataset

ADriverI integrates visionaction pairs using MLLM and diffusion models to autoregressively predict control signals and future scenes, enabling iterative autonomous driving and significantly improving performance.

Smoothing Grounding and Reasoning for MLLM-Powered GUI Agents with Query-Oriented Pivot Tasks

Published:3/1/2025

Multimodal Large Language ModelLLM-guided motion planningGraphical User Interface (GUI) AgentsQuery-Oriented ReasoningCoordinate Grounding and Reasoning Alignment

This work introduces query inference bridging coordinate grounding and action reasoning in MLLMpowered GUI agents, achieving superior performance with minimal data and enhanced reasoning by integrating semantic information.

A Survey on Generative Recommendation: Data, Model, and Tasks

Published:10/31/2025

Generative Recommendation SystemsLarge Language Model Fine-TuningDiffusion ModelsMultimodal Large Language ModelLLM-based Recommendation Systems

This survey reviews generative recommendation via a unified framework, analyzing data augmentation, model alignment, and task design, highlighting innovations in large language and diffusion models that enable knowledge integration, natural language understanding, and personalize

WebWatcher: Breaking New Frontiers of Vision-Language Deep Research Agent

Published:10/8/2025

Multimodal Large Language ModelVision-Language-Action ModelRL Training for Large Language ModelsComplex Information Retrieval BenchmarkVisual-Language Reasoning

WebWatcher is a multimodal deep research agent enhancing visuallanguage reasoning via synthetic trajectories and reinforcement learning, validated on the new BrowseCompVL benchmark for complex visualtext retrieval tasks, surpassing existing baselines.

Thinking in Space: How Multimodal Large Language Models See, Remember, and Recall Spaces

Published:12/19/2024

Multimodal Large Language ModelVisual-Spatial Intelligence BenchmarkSpatial ReasoningVideo Question AnsweringCognitive Map Generation

This work introduces VSIBench to evaluate multimodal large language models' spatial reasoning from videos, revealing emerging spatial awareness and local world models, with cognitive map generation enhancing spatial distance understanding beyond standard linguistic reasoning tec

Emerging Properties in Unified Multimodal Pretraining

Published:5/21/2025

Multimodal Large Language ModelLarge-Scale Multimodal PretrainingMultimodal Reasoning Capacity EnhancementMultimodal Generation and Understanding

BAGEL, a unified decoderonly multimodal model pretrained on massive interleaved data, shows emerging capabilities in complex reasoning, outperforming opensource models in generation and understanding, enabling advanced tasks like image manipulation and future frame prediction.

MQuant: Unleashing the Inference Potential of Multimodal Large Language Models via Static Quantization

Published:10/25/2025

Multimodal Large Language ModelStatic QuantizationPost-Training Quantization FrameworkModality-Specific QuantizationLLM Reasoning Capacity Enhancement

MQuant introduces a posttraining static quantization framework for multimodal large language models, reducing latency and outliers via modalityspecific quantization, flexible attention switching, and rotation suppression, boosting inference efficiency across major models.

KORE: Enhancing Knowledge Injection for Large Multimodal Models via Knowledge-Oriented Augmentations and Constraints

Published:10/22/2025

Multimodal Large Language ModelKnowledge Utilization in LLMsKnowledge Injection MethodsKnowledge Adaptation and RetentionCatastrophic Forgetting Mitigation

KORE enhances knowledge injection in large multimodal models by structured augmentations and constraints, preserving old knowledge via null space projection to mitigate forgetting and enable precise adaptation to new knowledge.

m3P: Towards Multimodal Multilingual Translation with Multimodal Prompt

Published:3/26/2024

Multimodal Large Language ModelMultimodal Multilingual TranslationMultimodal PromptMultilingual Multimodal Instruction DatasetLow-Resource Language Translation

m3P leverages multimodal prompts and visual context as a languageindependent representation to align 102 languages semantically, significantly improving translation quality, especially in lowresource settings.

LMAgent: A Large-scale Multimodal Agents Society for Multi-user Simulation

Published:12/12/2024

Multimodal Large Language ModelLarge-Scale Multi-Agent SystemsMulti-User Behavior SimulationE-Commerce Scenario SimulationSelf-Consistency Prompting

LMAgent develops a largescale multimodal agent society for realistic multiuser simulation in ecommerce, enhancing decisions via selfconsistency prompting and boosting efficiency with a smallworld fast memory, demonstrating humanlike behavior and herd effects.

010

Large Language Model Agent: A Survey on Methodology, Applications and Challenges

Published:3/27/2025

Survey on Large Language Model AgentsRL Training for Large Language ModelsLLM-guided motion planningLLM Reasoning Capacity EnhancementMultimodal Large Language Model

This survey systematically analyzes large language model agents’ architecture, collaboration, and evolution, unifying fragmented research and outlining evaluation, tools, and applications, to guide future advancements toward artificial general intelligence.

1 - 15 / 15

Papers