AI研究論文デイリー

LongVU: Spatiotemporal Adaptive Compression for Long Video-Language Understanding

TP-Eval: Tap Multimodal LLMs' Potential in Evaluation by Customizing Prompts

WorldSimBench: Towards Video Generation Models as World Simulators

M-RewardBench: Evaluating Reward Models in Multilingual Settings

Scaling Diffusion Language Models via Adaptation from Autoregressive Models

MIA-DPO: Multi-Image Augmented Direct Preference Optimization For Large Vision-Language Models

ARKit LabelMaker: A New Scale for Indoor 3D Scene Understanding

Scalable Ranked Preference Optimization for Text-to-Image Generation

Lightweight Neural App Control

MedINST: Meta Dataset of Biomedical Instructions

Steering Your Generalists: Improving Robotic Foundation Models via Value Guidance

LVSM: A Large View Synthesis Model with Minimal 3D Inductive Bias

DynamicCity: Large-Scale LiDAR Generation from Dynamic Scenes