持续更新中ing!!!
友情链接:大模型相关资料、基础技术和排行榜
大模型LLM论文目录
标题和时间 | 作者 | 来源 | 简介 |
---|---|---|---|
Artificial General Intelligence: Concept, State of the Art, and Future Prospects,2014 | Goertzel | Journal of Artificial General Intelligence | 14年的一篇AGI综述,里面探讨了AGI的定义、分类和评估方法等,作者貌似现在是AGI大会的编辑了hh |
Towards artificial general intelligence with hybrid Tianjic chip architecture,2020 | Pei jing | Nature | 2020年的一个讨论实现AGI硬件的论文,其实现了在同一芯片上同时运行MLP-like和SNN神经网络的硬件环境 |
AGI Brain II: The Upgraded Version with Increased Versatility Index,2021 | Mohammadreza Alidoust | AGI2021 | 1.提出一个AGI指标,2.用Mamdani模糊推理联想记忆代替原本的神经网络NN表示外显记忆 |
Training language models to follow instructions with human feedback,2022 | Long Ouyang等人 | OpenAI | InstructGPT,在大型语言模型的基础上引入人工引导和强化学习,大大提升模型性能 |
Yann Lecun: A Path Towards Autonomous Machine Intelligence 自主机器学习和AGI,2022 | Yann Lecun | Machine Learning | 提出了自主智能体的架构和训练范式,论文地址 |
GPT-4原论文详细解读(GPT-4 Technical Report),2023 | OpenAI | OpenAI | GPT-4,提出了多模态的大型语言模型,具备一定的常识和认知能力 |
ChatGLM,2023 | Aohan Zeng,Du等人 | International Conference on Learning Representations (ICLR) | ChatGLM,ChatGLM-6B结合模型量化技术,用户可以在消费级的显卡上进行本地部署(INT4 量化级别下最低只需 6GB 显存) |
LLaMA: Open and Efficient Foundation Language Models,2023 | Hugo Touvron | preprint | LLaMA 是 Meta AI 发布的包含 7-65B 参数规模的LLM,其中LLaMA-13B 仅以 1/10 规模的参数在多数的 benchmarks 上性能优于 GPT-3(175B)。开源。 |
A Survey of Large Language Models,2023 | Wayne Xin Zhao, | preprint | 大型语言模型综述,非常详细,格局打开! |
ChatDB: AUGMENTING LLMS WITH DATABASES AS THEIR SYMBOLIC MEMORY,2023 | Chenxu Hu | preprint | ChatDB清华团队针对大模型LLMs的长期记忆能力进行的改进,提出数据库与大模型结合开源 |
LONGNET: Scaling Transformers to1,000,000,000 Tokens,2023 | Jiayu Ding | preprint | LONGNET微软做的针对大模型的长文本学习,长期记忆进行的改进,开源 |
Focused Transformer: Contrastive Training for Context Scaling,2023 | Szymon Tworkowski | preprint | LongLlama谷歌DeepMind研究团队提出了一种注意力集中的transformer架构FOT |
Towards Benchmarking and Improving the Temporal Reasoning Capability of Large Language Models,2023 | 谭清宇,Hwee Tou Ng,邴立东 | ACL 2023 main conference | LLM理解时间变迁。达摩院联合NUS提出时间推理数据集以及时间强化的训练范式 |
UnIVAL: Unified Model for Image, Video, Audio and Language Tasks,2023 | Mustafa Shukor | preprint | UnIVAL,该算法不依赖于数据集大小或具有数十亿参数的大模型,仅仅具有约0.25B的参数量,而且将文本、图像、视频和音频这4个多模态任务统一到了一个模型中 |
Graph of Thoughts: Solving Elaborate Problems with Large Language Models,2023 | Besta Maciej | preprint | 思维图,将LLM生成的信息建模为任意图,其中信息单位是顶点,边代表顶点之间的依赖关系 |
The Rise and Potential of Large Language Model Based Agents: A Survey,2023 | Xi Zhi heng | preprint | Agent,综述 |
NExT-GPT: Any-to-Any Multimodal LLM,2023 | Wu Shengqiong | preprint | NExT-GPT,多模态大模型,实现任意模态之间的转换。NextGPT整体结构图、模型示意图和使用模型时示意图 |
Toolformer: Language Models Can Teach Themselves to Use Tools,2023 | Schick Timo | preprint | Toolsformer,GPT与各种工具结合 |
The Dawn of LMMs: Preliminary Explorations with GPT-4V(ision),2023 | Yang Zhengyuan | preprint | GPT-4V测评报告 |
EFFICIENT STREAMING LANGUAGE MODELS WITH ATTENTION SINKS,2023 | Xiao Guangxuan | preprint | 流式LLM,无限扩展LLM长度 |
Improving Image Generation with Better Captions,2023 | Betker James | Open AI | DaLLE3,作画大师接入chatgpt,论文中文版见这 |
Instruction Tuning for Large Language Models: A Survey,2023 | Zhang Linfeng | preprint | 思维链综述 |
RoleLLM: Benchmarking, Eliciting, and Enhancing Role-Playing Abilities of Large Language Models | Wang Zekun Moore | preprint | 角色扮演大模型 |
A Survey on Multimodal Large Language Models,2023 | Yin Chaoyou | preprint | 多模态大模型综述 |
Visual Instruction Tuning,2023 | Liu Haotian | preprint | 视觉大模型llava,通过视觉调优,支持基于图片的聊天 |
ChatGLM3,2023 | ZHIPU, Tinghua | web | ChatGLM3 |
AI Alignment: A Comprehensive Survey,2023 | Jiaming Ji | preprint | AI对齐技术综述,怎么让AI符合人类意图和价值观 |
RoboGen: Towards Unleashing Infinite Data for Automated Robot Learning via Generative Simulation,2023 | Yufei Wang | preprint | 具身智能代表性工作 |
A Comprehensive Overview of Large Language Models, 2023 | Naveed Humza | arXiv | 大模型的全面回顾,看透大模型 |