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。
DeepSeek-AI, Daya Guo, Dejian Yang +197 authors
DeepSeek-R1-Zero and DeepSeek-R1 utilize reinforcement learning and multi-stage training to enhance reasoning capabilities, with DeepSeek-R1 achieving performance comparable to OpenAI-o1-1217.
50 篇论文 · 按点赞排序
MiniMax, Aonian Li, Bangwei Gong +87 authors
The MiniMax-01 series, including MiniMax-Text-01 and MiniMax-VL-01, offer superior long-context processing and match state-of-the-art model performance with longer context windows through lightning attention, Mixture of Experts (MoE), and efficient parallel strategies.
Xinyu Guan, Li Lyna Zhang, Yifei Liu +5 authors
rStar-Math enhances small language models' math reasoning capabilities through Monte Carlo Tree Search and self-evolution, achieving state-of-the-art performance on various benchmarks without distillation from larger models.
Kimi Team, Angang Du, Bofei Gao +91 authors
A multi-modal LLM trained with reinforcement learning achieves state-of-the-art reasoning performance across various benchmarks by utilizing long context scaling and effective policy optimization methods.
Tianzhe Chu, Yuexiang Zhai, Jihan Yang +6 authors
Reinforcement learning demonstrates superior generalization compared to supervised fine-tuning across textual and visual domains, while supervised fine-tuning stabilizes the model output format essential for RL performance.
Kuang-Huei Lee, Ian Fischer, Yueh-Hua Wu +4 authors
Mind Evolution, an evolutionary search strategy using a language model, outperforms other inference methods in natural language planning tasks by generating, recombining, and refining candidate responses.
Wenqi Zhang, Hang Zhang, Xin Li +6 authors
A new high-quality instructional video-based corpus improves VLM pretraining through more coherent context and richer knowledge, enhancing performance in knowledge-intensive tasks.
Siyu Yuan, Zehui Chen, Zhiheng Xi +3 authors
Agent-R, an iterative self-training framework using Monte Carlo Tree Search, enhances large language models' ability to recover from errors and correct actions in real-time within interactive environments.
Xiaoxi Li, Guanting Dong, Jiajie Jin +5 authors
Search-o1 enhances large reasoning models with an agentic retrieval-augmented generation mechanism and a Reason-in-Documents module to improve performance on complex reasoning tasks.
Jian Hu
REINFORCE++ is an enhanced reinforcement learning algorithm that combines PPO techniques without a critic network, offering simplicity, stability, and efficiency.
Zhenru Zhang, Chujie Zheng, Yangzhen Wu +6 authors
Process Reward Models improve performance in mathematical reasoning by integrating Monte Carlo estimation with LLM-as-a-judge, addressing biases in Best-of-N evaluation and step-wise error identification.
Yiwen Huang, Aaron Gokaslan, Volodymyr Kuleshov +1 authors
A new regularization approach and simplified architecture for GANs, named R3GAN, surpasses existing methods including StyleGAN2 and diffusion models across multiple datasets.
Violet Xiang, Charlie Snell, Kanishk Gandhi +11 authors
The Meta Chain-of-Thought framework extends traditional CoT by explicitly modeling reasoning, enabling more powerful and human-like reasoning in LLMs through process supervision, synthetic data generation, and reinforcement learning.
Samuel Schmidgall, Yusheng Su, Ze Wang +6 authors
Agent Laboratory, an LLM-based framework, accelerates scientific research by automating the entire process from idea to report, resulting in high-quality outputs with reduced expenses.
Boqiang Zhang, Kehan Li, Zesen Cheng +12 authors
A vision-centric multimodal foundation model, VideoLLaMA3, improves image and video understanding through strategic training stages and adaptive vision token encoding.
Yifan Zhang, Yifeng Liu, Huizhuo Yuan +4 authors
T6, a new Transformer architecture using Tensor Product Attention, improves language model performance and memory efficiency, allowing for longer sequence processing.
Yue Liu, Hongcheng Gao, Shengfang Zhai +7 authors
GuardReasoner enhances LLM safety through enhanced reasoning capabilities using specialized training datasets and techniques, outperforming existing models on guardrail tasks.
Qiushi Sun, Kanzhi Cheng, Zichen Ding +12 authors
OS-Genesis is a novel GUI data synthesis pipeline that enhances the training of GUI agents by reversing the trajectory collection process to improve data quality and diversity.
NVIDIA, Niket Agarwal, Arslan Ali +75 authors
The Cosmos World Foundation Model Platform provides a framework for developing customized world models for Physical AI by leveraging video curation, pre-trained models, and tokenizers.
Yilun Zhao, Lujing Xie, Haowei Zhang +16 authors
A comprehensive benchmark for evaluating foundation models in video understanding includes domain-specific questions and expert evaluations, highlighting the gap between current models and human expertise.
Long Phan, Alice Gatti, Ziwen Han +1155 authors
HLE is a challenging multi-modal benchmark that highlights the limitations of current LLMs in closed-ended academic questions.
Zhenran Xu, Longyue Wang, Jifang Wang +7 authors
FilmAgent, a multi-agent collaborative framework based on LLMs, automates film production processes including scriptwriting, cinematography, and actor positioning, outperforming other models in human evaluations.
Soyeong Jeong, Kangsan Kim, Jinheon Baek +1 authors
VideoRAG dynamically retrieves relevant videos based on queries and utilizes both visual and textual information for output generation, outperforming existing methods in response accuracy.
An Yang, Bowen Yu, Chengyuan Li +25 authors
The Qwen2.5-1M series models extend context length to 1 million tokens with enhanced long-context capabilities, employing techniques like long data synthesis and progressive pre-training, and are supported by an open-source inference framework with sparse attention and kernel optimizations.
Nanye Ma, Shangyuan Tong, Haolin Jia +8 authors
Inference-time scaling of diffusion models improves sample quality beyond increasing denoising steps through better noise selection and verification.
Zhengyang Tang, Ziniu Li, Zhenyang Xiao +8 authors
SCRIT, a self-evolving critique framework, enhances LLMs' critique capabilities using synthetic data and self-validation, achieving significant improvements in critique-correction and error identification benchmarks.
Alsu Sagirova, Yuri Kuratov, Mikhail Burtsev
Shared Recurrent Memory Transformer (SRMT) enhances cooperation in multi-agent reinforcement learning through implicit information exchange, outperforming baselines in navigation tasks and generalizing to unseen scenarios.
Zihan Qiu, Zeyu Huang, Bo Zheng +7 authors
Calculating Load-Balancing Loss using a global-batch instead of a micro-batch improves performance and domain specialization in MoE-based large language models.
Omkar Thawakar, Dinura Dissanayake, Ketan More +12 authors
A framework for evaluating and improving step-by-step visual reasoning in large language models using a specialized benchmark and a novel multimodal model trained with curriculum learning.
Yujia Qin, Yining Ye, Junjie Fang +32 authors
UI-TARS, a native GUI agent model using screenshots as input, outperforms commercial models in various benchmarks through enhanced perception, unified action modeling, system-2 reasoning, and iterative training with reflective online traces.
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