AI研究論文每日精選

MM1.5: Methods, Analysis & Insights from Multimodal LLM Fine-tuning

Ruler: A Model-Agnostic Method to Control Generated Length for Large Language Models

Hyper-Connections

DiaSynth -- Synthetic Dialogue Generation Framework

Cottention: Linear Transformers With Cosine Attention

UniAff: A Unified Representation of Affordances for Tool Usage and Articulation with Vision-Language Models

Scaling Proprioceptive-Visual Learning with Heterogeneous Pre-trained Transformers

Image Copy Detection for Diffusion Models

Can Models Learn Skill Composition from Examples?

Coffee-Gym: An Environment for Evaluating and Improving Natural Language Feedback on Erroneous Code

Visual Question Decomposition on Multimodal Large Language Models

IDEAW: Robust Neural Audio Watermarking with Invertible Dual-Embedding