Explore · 每周精选
发现最受关注的研究论文,追踪研究趋势,订阅感兴趣的期刊与关键词。
50 篇论文 · 按点赞排序
Yuyang Zhao, Yicheng Pan, Qiyuan He +6 authors
SANA-Streaming enables real-time high-resolution video-to-video editing through a hybrid diffusion transformer architecture, cycle-reverse regularization, and efficient system co-design optimized for consumer GPUs.
Tao Zou, Yichen He, Tian Qiu +2 authors
A reinforcement-learning-based framework called TaskMem is introduced to dynamically determine what information to store in long-term memory for multimodal agents, improving performance on streaming video benchmarks.
Lin Fu, Zheyuan Yang, Yang Wang +3 authors
VideoKR presents a large-scale video reasoning dataset and benchmark designed to enhance knowledge-intensive video understanding through expert-domain content and human-in-the-loop example generation.
Tianhe Wu, Kun Yan, Zikai Zhou +21 authors
Few-step distillation for visual generative models benefits from systematic investigation of training recipes beyond just distillation objectives, leading to improved student performance through optimized data composition, teacher guidance, and task mixture.
Zhuoyun Yu, Xin Xie, Wuguannan Yao +4 authors
Step-level skill adaptation framework with explicit failure attribution improves training-free skill maintenance for LLM agents in interactive tasks.
Ke Lei, Yu Zhang, Changhao Pan +4 authors
SwanSphere presents a unified streaming framework for high-fidelity spatial audio generation from panoramic videos and text prompts using causal autoregressive diffusion transformers and multimodal learning strategies.
Sicheng Feng, Zigeng Chen, Gongfan Fang +2 authors
Diffusion large language models combined with mixture-of-experts architectures face a mismatch between block parallel decoding and token-level expert selection, which dMoE addresses by aggregating token-level distributions into block-level routing to reduce activated experts and improve efficiency.
Peiwen Sun, Xudong Lu, Huadai Liu +9 authors
X-Stream introduces the first benchmark for multi-stream streaming understanding, revealing significant limitations of current MLLMs in handling concurrent streams.
Haodi Lei, Yafy Li, Haoran Zhang +8 authors
Speculative decoding uses a lightweight draft model to accelerate large language model inference, but supervised fine-tuning plateaus due to offline-to-inference mismatch, which is addressed through on-policy distillation with target-assisted rollouts and error replay.
Xiangdong Zhang, Debing Zhang, Shaofeng Zhang +3 authors
Next Implicit Token Prediction enhances language model training by adding dense continuous supervision in representation space, improving generalization and performance across model sizes with minimal computational overhead.
Changhao Pan, Rui Yang, Han Wang +12 authors
Swanbench-Speech addresses the lack of comprehensive long-form speech evaluation by providing a benchmark with diverse scenarios, multi-dimensional metrics, and insights into model limitations.
Zhangchen Xu, Junda Chen, Yue Huang +16 authors
AutoLab benchmark evaluates long-horizon iterative optimization capabilities of frontier models across diverse domains, revealing that persistent iteration and time awareness are more critical than initial performance quality.
Yifei Li, Pengyiang Liu, Yuhang Zang +4 authors
OVO-S-Bench presents a comprehensive benchmark for evaluating streaming spatial intelligence in multimodal language models through human-annotated questions spanning multiple abstraction levels.
Liyang Li, Muzhi Zhu, Zhiyue Zhao +5 authors
Target Viewpoint Reproduction task challenges foundation models to actively adjust 3D viewpoints to match target images, revealing limitations in visual history processing and embodied movement mapping, with a unified post-training framework improving success rates through various training methods.
Yucheng Zhou, Wei Tao, Yiwen Guo +1 authors
Controlled concrete reasoning combines visual simulation with abstract reasoning through a training method that uses privileged future information to improve prediction accuracy and robustness.
Zhen Yang, Xiaogang Xu, Wen Wang +3 authors
StreamMA enables efficient multi-agent reasoning by streaming intermediate results and leveraging reliable early steps to improve both latency and effectiveness across various reasoning tasks.
Ali Behrouz, Farnoosh Hashemi, Vahab Mirrokni
Deep learning models with sleep and dreaming paradigms enable continual learning through memory consolidation and self-improvement phases.
Yuxuan Bian, Zeyue Xue, Songchun Zhang +9 authors
Echo Infinity enables real-time infinite video generation using learnable evolving memory and unified relative RoPE to overcome limitations in existing autoregressive methods.
Junhao Cheng, Liang Hou, Tianxiong Zhong +4 authors
Video generation models combined with vision-language models acting as test-time teachers through differentiable rewards achieve superior video reasoning performance.
Yifeng Liu, Shiyuan Zhang, Yifan Zhang +1 authors
A self-distilled policy-gradient framework combines on-policy self-distillation with verifier advantages and KL regularization to improve reinforcement learning stability and performance.
北京市昌平区探索星信息技术及软件开发工作室
京ICP备2026059466号