Trends · 研究趋势
数据来自 Hugging Face 论文的 AI 提取关键词,按月统计研究方向的增长与热度。
Qwen, An Yang, Baosong Yang +39 authors
Qwen2.5, an enhanced series of large language models, demonstrates superior performance across various benchmarks and use cases through extensive pre-training and advanced post-training techniques.
Wenyi Hong, Wenmeng Yu, Xiaotao Gu +74 authors
A vision-language model (VLM) named GLM-4.1V-Thinking, developed with a reasoning-centric training framework, achieves state-of-the-art performance across various tasks, including STEM problem solving, video understanding, and long document understanding, outperforming larger models on many benchmarks.
Deepak Nathani, Lovish Madaan, Nicholas Roberts +14 authors
A new framework and benchmark, MLGym and MLGym-Bench, is introduced for evaluating and developing LLM agents across diverse AI research tasks using reinforcement learning.
Haozhe Wang, Haoran Que, Qixin Xu +9 authors
REER, a new paradigm for deep reasoning, uses reverse engineering to discover step-by-step reasoning processes, enabling a model to perform competitively on open-ended tasks.
Shen Nie, Fengqi Zhu, Zebin You +7 authors
LLaDA, a diffusion model trained from scratch, outperforms autoregressive models in benchmarks and demonstrates strong instruction-following capabilities, challenging the dominance of ARMs in LLMs.
Matt Deitke, Christopher Clark, Sangho Lee +48 authors
Molmo, a new family of open-weight VLMs, excels through a high-quality human-annotated image caption dataset and a diverse fine-tuning dataset, outperforming both proprietary and open models in benchmark and human evaluations.
Elisei Rykov, Kseniia Petrushina, Maksim Savkin +6 authors
PsiloQA, a multilingual dataset with span-level hallucinations, enhances hallucination detection in large language models across 14 languages using an automated pipeline and encoder-based models.
Zhuofeng Li, Haoxiang Zhang, Seungju Han +6 authors
AgentFlow, a trainable agentic framework with in-the-flow optimization, enhances reasoning in large language models by coordinating specialized modules and outperforms top baselines across various tasks.
Yibin Wang, Zhimin Li, Yuhang Zang +4 authors
A unified multimodal reward model that incorporates long chains of thought reasoning improves reliability and accuracy in vision reward tasks through exploration-driven reinforcement learning.
Philippe Laban, Alexander R. Fabbri, Caiming Xiong +1 authors
The SummHay task evaluates LLMs and RAG systems on summarizing long-context documents by identifying relevant insights and citing sources, revealing challenges even for systems with document relevance signals.
Chujie Zheng, Zhenru Zhang, Beichen Zhang +6 authors
ProcessBench evaluates models' ability to identify errors in mathematical reasoning steps, showing that existing process reward models struggle with difficult problems and underperform compared to critic models and a fine-tuned PRM.
Pooyan Rahmanzadehgervi, Logan Bolton, Mohammad Reza Taesiri +1 authors
State-of-the-art large language models with vision capabilities perform poorly on simple visual tasks, indicating limitations in their visual understanding.
Xiusi Chen, Gaotang Li, Ziqi Wang +9 authors
Reasoning Reward Models (ReasRMs) enhance reward modeling for large language models by integrating reasoning tasks, improving interpretability and performance.
Runqi Qiao, Qiuna Tan, Guanting Dong +15 authors
WE-MATH benchmark evaluates Large Multimodal Models in visual mathematical reasoning, focusing on knowledge acquisition and generalization principles beyond end-to-end performance.
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.
Ziyu Liu, Tao Chu, Yuhang Zang +8 authors
A new benchmark (MMDU) and instruction tuning dataset (MMDU-45k) are introduced to evaluate and enhance large vision-language models' performance in multi-turn and multi-image conversations, highlighting the need for improved conversational instruction tuning data.
Lin Long, Yichen He, Wentao Ye +5 authors
M3-Agent, a multimodal agent with long-term memory, performs multi-turn reasoning and outperforms baselines on a new long-video question answering benchmark.
Zhaorun Chen, Yichao Du, Zichen Wen +16 authors
MJ-Bench evaluates multimodal judges across alignment, safety, image quality, and bias for text-to-image models and confirms close-source VLMs generally provide better feedback.
Zekun Wang, King Zhu, Chunpu Xu +14 authors
MIO, a novel foundation model, achieves competitive performance in multimodal tasks through an end-to-end, autoregressive approach using causal multimodal modeling.
Shaolei Zhang, Qingkai Fang, Zhe Yang +1 authors
LLaVA-Mini, a unified large multimodal model, significantly reduces computational cost and latency by compressing vision tokens to a single token before feeding them to the LLM backbone, outperforming previous models.
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