Ежедневные Исследовательские Статьи по ИИ

Arctic-SnowCoder: Demystifying High-Quality Data in Code Pretraining

LongLLaVA: Scaling Multi-modal LLMs to 1000 Images Efficiently via Hybrid Architecture

LongCite: Enabling LLMs to Generate Fine-grained Citations in Long-context QA

Affordance-based Robot Manipulation with Flow Matching

FastVoiceGrad: One-step Diffusion-Based Voice Conversion with Adversarial Conditional Diffusion Distillation

Loopy: Taming Audio-Driven Portrait Avatar with Long-Term Motion Dependency

Political DEBATE: Efficient Zero-shot and Few-shot Classifiers for Political Text

MMMU-Pro: A More Robust Multi-discipline Multimodal Understanding Benchmark