AI研究論文每日精選

Hogwild! Inference: Parallel LLM Generation via Concurrent Attention

Leanabell-Prover: Posttraining Scaling in Formal Reasoning

Tuning-Free Image Editing with Fidelity and Editability via Unified Latent Diffusion Model

Unified World Models: Coupling Video and Action Diffusion for Pretraining on Large Robotic Datasets

Generative Evaluation of Complex Reasoning in Large Language Models

V-MAGE: A Game Evaluation Framework for Assessing Visual-Centric Capabilities in Multimodal Large Language Models

CrossWordBench: Evaluating the Reasoning Capabilities of LLMs and LVLMs with Controllable Puzzle Generation

Accelerate Parallelizable Reasoning via Parallel Decoding within One Sequence

An Empirical Study of GPT-4o Image Generation Capabilities

ProtoGCD: Unified and Unbiased Prototype Learning for Generalized Category Discovery

Less-to-More Generalization: Unlocking More Controllability by In-Context Generation

COIG-P: A High-Quality and Large-Scale Chinese Preference Dataset for Alignment with Human Values

Efficient Reinforcement Finetuning via Adaptive Curriculum Learning

HybriMoE: Hybrid CPU-GPU Scheduling and Cache Management for Efficient MoE Inference

Skywork R1V: Pioneering Multimodal Reasoning with Chain-of-Thought

HiFlow: Training-free High-Resolution Image Generation with Flow-Aligned Guidance

OmniSVG: A Unified Scalable Vector Graphics Generation Model