AI 연구 논문 데일리

TabSTAR: A Foundation Tabular Model With Semantically Target-Aware Representations

QwenLong-L1: Towards Long-Context Large Reasoning Models with Reinforcement Learning

Distilling LLM Agent into Small Models with Retrieval and Code Tools

Quartet: Native FP4 Training Can Be Optimal for Large Language Models

Reasoning Model is Stubborn: Diagnosing Instruction Overriding in Reasoning Models

One RL to See Them All: Visual Triple Unified Reinforcement Learning

PhyX: Does Your Model Have the "Wits" for Physical Reasoning?

QwenLong-CPRS: Towards infty-LLMs with Dynamic Context Optimization

Scaling Image and Video Generation via Test-Time Evolutionary Search

MOOSE-Chem3: Toward Experiment-Guided Hypothesis Ranking via Simulated Experimental Feedback

Model Already Knows the Best Noise: Bayesian Active Noise Selection via Attention in Video Diffusion Model

VeriThinker: Learning to Verify Makes Reasoning Model Efficient

Diffusion Classifiers Understand Compositionality, but Conditions Apply

AudioTrust: Benchmarking the Multifaceted Trustworthiness of Audio Large Language Models

Direct3D-S2: Gigascale 3D Generation Made Easy with Spatial Sparse Attention

Position of Uncertainty: A Cross-Linguistic Study of Positional Bias in Large Language Models

s3: You Don't Need That Much Data to Train a Search Agent via RL

FullFront: Benchmarking MLLMs Across the Full Front-End Engineering Workflow

Thought-Augmented Policy Optimization: Bridging External Guidance and Internal Capabilities

Time-R1: Towards Comprehensive Temporal Reasoning in LLMs

Teaching with Lies: Curriculum DPO on Synthetic Negatives for Hallucination Detection

Clear Nights Ahead: Towards Multi-Weather Nighttime Image Restoration

Speechless: Speech Instruction Training Without Speech for Low Resource Languages

RBench-V: A Primary Assessment for Visual Reasoning Models with Multi-modal Outputs

Teaching Large Language Models to Maintain Contextual Faithfulness via Synthetic Tasks and Reinforcement Learning

Trinity-RFT: A General-Purpose and Unified Framework for Reinforcement Fine-Tuning of Large Language Models

ScanBot: Towards Intelligent Surface Scanning in Embodied Robotic Systems

Are Vision-Language Models Safe in the Wild? A Meme-Based Benchmark Study

Synthetic Data RL: Task Definition Is All You Need

DanceTogether! Identity-Preserving Multi-Person Interactive Video Generation

RePrompt: Reasoning-Augmented Reprompting for Text-to-Image Generation via Reinforcement Learning

Transformer Copilot: Learning from The Mistake Log in LLM Fine-tuning

On the Design of KL-Regularized Policy Gradient Algorithms for LLM Reasoning

Interactive Post-Training for Vision-Language-Action Models

ReflAct: World-Grounded Decision Making in LLM Agents via Goal-State Reflection

Large Language Models Implicitly Learn to See and Hear Just By Reading

Value-Guided Search for Efficient Chain-of-Thought Reasoning

Not All Models Suit Expert Offloading: On Local Routing Consistency of Mixture-of-Expert Models

Keep Security! Benchmarking Security Policy Preservation in Large Language Model Contexts Against Indirect Attacks in Question Answering

Revisiting Residual Connections: Orthogonal Updates for Stable and Efficient Deep Networks

FREESON: Retriever-Free Retrieval-Augmented Reasoning via Corpus-Traversing MCTS

Augmenting LLM Reasoning with Dynamic Notes Writing for Complex QA

NOVER: Incentive Training for Language Models via Verifier-Free Reinforcement Learning

TIME: A Multi-level Benchmark for Temporal Reasoning of LLMs in Real-World Scenarios

NileChat: Towards Linguistically Diverse and Culturally Aware LLMs for Local Communities

FuxiMT: Sparsifying Large Language Models for Chinese-Centric Multilingual Machine Translation

Universal Biological Sequence Reranking for Improved De Novo Peptide Sequencing