AI研究論文デイリー

Qwen2-Audio Technical Report

NeedleBench: Can LLMs Do Retrieval and Reasoning in 1 Million Context Window?

Scaling Diffusion Transformers to 16 Billion Parameters

Ref-AVS: Refer and Segment Objects in Audio-Visual Scenes

Sibyl: Simple yet Effective Agent Framework for Complex Real-world Reasoning

VLMEvalKit: An Open-Source Toolkit for Evaluating Large Multi-Modality Models

DreamCatalyst: Fast and High-Quality 3D Editing via Controlling Editability and Identity Preservation

Animate3D: Animating Any 3D Model with Multi-view Video Diffusion

Efficient Training with Denoised Neural Weights

FIRE: A Dataset for Feedback Integration and Refinement Evaluation of Multimodal Models

YouTube-SL-25: A Large-Scale, Open-Domain Multilingual Sign Language Parallel Corpus

EfficientQAT: Efficient Quantization-Aware Training for Large Language Models

From GaLore to WeLore: How Low-Rank Weights Non-uniformly Emerge from Low-Rank Gradients

OmniBind: Large-scale Omni Multimodal Representation via Binding Spaces

Grasping Diverse Objects with Simulated Humanoids

Vibravox: A Dataset of French Speech Captured with Body-conduction Audio Sensors

Data-Juicer Sandbox: A Comprehensive Suite for Multimodal Data-Model Co-development

Click-Gaussian: Interactive Segmentation to Any 3D Gaussians

Uncertainty is Fragile: Manipulating Uncertainty in Large Language Models