AI 연구 논문 데일리

MLGym: A New Framework and Benchmark for Advancing AI Research Agents

SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features

SuperGPQA: Scaling LLM Evaluation across 285 Graduate Disciplines

How Much Knowledge Can You Pack into a LoRA Adapter without Harming LLM?

S*: Test Time Scaling for Code Generation

Logic-RL: Unleashing LLM Reasoning with Rule-Based Reinforcement Learning

Discovering highly efficient low-weight quantum error-correcting codes with reinforcement learning

LongWriter-V: Enabling Ultra-Long and High-Fidelity Generation in Vision-Language Models

S^2R: Teaching LLMs to Self-verify and Self-correct via Reinforcement Learning

Does Time Have Its Place? Temporal Heads: Where Language Models Recall Time-specific Information

PC-Agent: A Hierarchical Multi-Agent Collaboration Framework for Complex Task Automation on PC

Dynamic Concepts Personalization from Single Videos

Scaling Text-Rich Image Understanding via Code-Guided Synthetic Multimodal Data Generation

How to Get Your LLM to Generate Challenging Problems for Evaluation

NAVIG: Natural Language-guided Analysis with Vision Language Models for Image Geo-localization

RelaCtrl: Relevance-Guided Efficient Control for Diffusion Transformers

From RAG to Memory: Non-Parametric Continual Learning for Large Language Models

AlphaMaze: Enhancing Large Language Models' Spatial Intelligence via GRPO

LServe: Efficient Long-sequence LLM Serving with Unified Sparse Attention

Enhancing Cognition and Explainability of Multimodal Foundation Models with Self-Synthesized Data

Generating Skyline Datasets for Data Science Models

CLIPPER: Compression enables long-context synthetic data generation

LLM-based User Profile Management for Recommender System

Multimodal RewardBench: Holistic Evaluation of Reward Models for Vision Language Models

How Much Do LLMs Hallucinate across Languages? On Multilingual Estimation of LLM Hallucination in the Wild

Unstructured Evidence Attribution for Long Context Query Focused Summarization

Geolocation with Real Human Gameplay Data: A Large-Scale Dataset and Human-Like Reasoning Framework

Generating π-Functional Molecules Using STGG+ with Active Learning

Symmetrical Visual Contrastive Optimization: Aligning Vision-Language Models with Minimal Contrastive Images