AI研究論文每日精選

ScienceBoard: Evaluating Multimodal Autonomous Agents in Realistic Scientific Workflows

Paper2Poster: Towards Multimodal Poster Automation from Scientific Papers

MME-Reasoning: A Comprehensive Benchmark for Logical Reasoning in MLLMs

OmniConsistency: Learning Style-Agnostic Consistency from Paired Stylization Data

SynLogic: Synthesizing Verifiable Reasoning Data at Scale for Learning Logical Reasoning and Beyond

Exploring the Latent Capacity of LLMs for One-Step Text Generation

OpenS2V-Nexus: A Detailed Benchmark and Million-Scale Dataset for Subject-to-Video Generation

Don't Overthink it. Preferring Shorter Thinking Chains for Improved LLM Reasoning

MMMR: Benchmarking Massive Multi-Modal Reasoning Tasks

Guided by Gut: Efficient Test-Time Scaling with Reinforced Intrinsic Confidence

VerIPO: Cultivating Long Reasoning in Video-LLMs via Verifier-Gudied Iterative Policy Optimization

Sparse VideoGen2: Accelerate Video Generation with Sparse Attention via Semantic-Aware Permutation

MME-VideoOCR: Evaluating OCR-Based Capabilities of Multimodal LLMs in Video Scenarios

UI-Genie: A Self-Improving Approach for Iteratively Boosting MLLM-based Mobile GUI Agents

GraLoRA: Granular Low-Rank Adaptation for Parameter-Efficient Fine-Tuning

Video-Holmes: Can MLLM Think Like Holmes for Complex Video Reasoning?

SweEval: Do LLMs Really Swear? A Safety Benchmark for Testing Limits for Enterprise Use

rStar-Coder: Scaling Competitive Code Reasoning with a Large-Scale Verified Dataset

Reinforcing General Reasoning without Verifiers

MetaMind: Modeling Human Social Thoughts with Metacognitive Multi-Agent Systems

Code Graph Model (CGM): A Graph-Integrated Large Language Model for Repository-Level Software Engineering Tasks

HoliTom: Holistic Token Merging for Fast Video Large Language Models

MotionPro: A Precise Motion Controller for Image-to-Video Generation

Beyond Distillation: Pushing the Limits of Medical LLM Reasoning with Minimalist Rule-Based RL

How does Alignment Enhance LLMs' Multilingual Capabilities? A Language Neurons Perspective

NOVA: A Benchmark for Anomaly Localization and Clinical Reasoning in Brain MRI

ImgEdit: A Unified Image Editing Dataset and Benchmark

Frame In-N-Out: Unbounded Controllable Image-to-Video Generation

DetailFlow: 1D Coarse-to-Fine Autoregressive Image Generation via Next-Detail Prediction

Active-O3: Empowering Multimodal Large Language Models with Active Perception via GRPO

Beyond Prompt Engineering: Robust Behavior Control in LLMs via Steering Target Atoms

FinTagging: An LLM-ready Benchmark for Extracting and Structuring Financial Information

ViewSpatial-Bench: Evaluating Multi-perspective Spatial Localization in Vision-Language Models

Thinker: Learning to Think Fast and Slow

Rendering-Aware Reinforcement Learning for Vector Graphics Generation

VisualToolAgent (VisTA): A Reinforcement Learning Framework for Visual Tool Selection

Adversarial Attacks against Closed-Source MLLMs via Feature Optimal Alignment

SeePhys: Does Seeing Help Thinking? -- Benchmarking Vision-Based Physics Reasoning

MMMG: a Comprehensive and Reliable Evaluation Suite for Multitask Multimodal Generation

MMPerspective: Do MLLMs Understand Perspective? A Comprehensive Benchmark for Perspective Perception, Reasoning, and Robustness

Alita: Generalist Agent Enabling Scalable Agentic Reasoning with Minimal Predefinition and Maximal Self-Evolution

VideoGameBench: Can Vision-Language Models complete popular video games?

Scaling External Knowledge Input Beyond Context Windows of LLMs via Multi-Agent Collaboration

Walk Before You Run! Concise LLM Reasoning via Reinforcement Learning

Minute-Long Videos with Dual Parallelisms

Beyond Markovian: Reflective Exploration via Bayes-Adaptive RL for LLM Reasoning

Can Compressed LLMs Truly Act? An Empirical Evaluation of Agentic Capabilities in LLM Compression

BiomedSQL: Text-to-SQL for Scientific Reasoning on Biomedical Knowledge Bases

R1-Searcher++: Incentivizing the Dynamic Knowledge Acquisition of LLMs via Reinforcement Learning

Search and Refine During Think: Autonomous Retrieval-Augmented Reasoning of LLMs

Sci-Fi: Symmetric Constraint for Frame Inbetweening

SoloSpeech: Enhancing Intelligibility and Quality in Target Speech Extraction through a Cascaded Generative Pipeline

MLLMs are Deeply Affected by Modality Bias

Spatial Knowledge Graph-Guided Multimodal Synthesis

Inverse Virtual Try-On: Generating Multi-Category Product-Style Images from Clothed Individuals

VLM-3R: Vision-Language Models Augmented with Instruction-Aligned 3D Reconstruction

Capability-Based Scaling Laws for LLM Red-Teaming

DFIR-Metric: A Benchmark Dataset for Evaluating Large Language Models in Digital Forensics and Incident Response

Modality Curation: Building Universal Embeddings for Advanced Multimodal Information Retrieval

ComfyMind: Toward General-Purpose Generation via Tree-Based Planning and Reactive Feedback

AdInject: Real-World Black-Box Attacks on Web Agents via Advertising Delivery

SATORI-R1: Incentivizing Multimodal Reasoning with Spatial Grounding and Verifiable Rewards

PreMoe: Lightening MoEs on Constrained Memory by Expert Pruning and Retrieval

R1-ShareVL: Incentivizing Reasoning Capability of Multimodal Large Language Models via Share-GRPO

Absolute Coordinates Make Motion Generation Easy

CoreMatching: A Co-adaptive Sparse Inference Framework with Token and Neuron Pruning for Comprehensive Acceleration of Vision-Language Models

Explaining Sources of Uncertainty in Automated Fact-Checking

Tropical Attention: Neural Algorithmic Reasoning for Combinatorial Algorithms

Improving Chemical Understanding of LLMs via SMILES Parsing

Do RAG Systems Suffer From Positional Bias?

Vision Transformers with Self-Distilled Registers

Ankh3: Multi-Task Pretraining with Sequence Denoising and Completion Enhances Protein Representations

Beyond Simple Concatenation: Fairly Assessing PLM Architectures for Multi-Chain Protein-Protein Interactions Prediction

An Explainable Diagnostic Framework for Neurodegenerative Dementias via Reinforcement-Optimized LLM Reasoning