최신 연구 논문

EarthMind: Towards Multi-Granular and Multi-Sensor Earth Observation with Large Multimodal Models

SRPO: Enhancing Multimodal LLM Reasoning via Reflection-Aware Reinforcement Learning

ShapeLLM-Omni: A Native Multimodal LLM for 3D Generation and Understanding

LoHoVLA: A Unified Vision-Language-Action Model for Long-Horizon Embodied Tasks

Temporal In-Context Fine-Tuning for Versatile Control of Video Diffusion Models

Jigsaw-R1: A Study of Rule-based Visual Reinforcement Learning with Jigsaw Puzzles

SmolVLA: A Vision-Language-Action Model for Affordable and Efficient Robotics

Taming LLMs by Scaling Learning Rates with Gradient Grouping

REASONING GYM: Reasoning Environments for Reinforcement Learning with Verifiable Rewards

Beyond the 80/20 Rule: High-Entropy Minority Tokens Drive Effective Reinforcement Learning for LLM Reasoning