视觉多模态话题 自然语言话题 生命科学话题 基础研究话题
1)标题:Meta|data2vec: A General Framework for Self-supervised Learning in Speech, Vision and Language(data2vec:语音、视觉和语言自监督学习的通用框架)
https://arxiv.org/pdf/2212.07525.pdf
https://hub.baai.ac.cn/view/14313
2)标题:谷歌 | Scaling Autoregressive Models for Content-Rich Text-to-Image Generation(用于内容丰富的文本到图像生成的缩放自回归模型)
https://arxiv.org/pdf/2206.10789.pdf
https://hub.baai.ac.cn/view/18292
3)标题: Meta | Scaling Language-Image Pre-training via Masking(通过掩码进行语言-图像预训练的缩放)
https://arxiv.org/pdf/2212.00794.pdf
https://hub.baai.ac.cn/view/22214
4)标题:谷歌等 | RT-1: Robotics Transformer for Real-World Control at Scale(RT-1: 用于真实世界大规模控制的机器人Transformer)
https://arxiv.org/pdf/2212.06817.pdf
https://hub.baai.ac.cn/view/22538
5)标题:Deepmind、牛津等 | Pushing the limits of self-supervised ResNets: Can we outperform supervised learning without labels on ImageNet?(突破自监督ResNet的极限:我们能否在无标签ImageNet上超越监督学习?)
https://arxiv.org/pdf/2201.05119v1.pdf
https://hub.baai.ac.cn/view/14183
6)标题:谷歌 | PaLI: A Jointly-Scaled Multilingual Language-Image Model(一种联合扩展的多语种语言图像模型)
https://arxiv.org/abs/2209.06794
https://hub.baai.ac.cn/view/20656
7)标题:斯坦福、Salesforce|MaskViT: Masked Visual Pre-Training for Video Prediction(MaskViT:遮蔽视觉预训练用于视频预测)
https://arxiv.org/pdf/2206.11894v1.pdf
https://hub.baai.ac.cn/view/18528
8)标题:百度 | 通过原型进行提示:基于原型的预训练视觉语言模型的Prompt 学习
https://arxiv.org/pdf/2210.10841.pdf
https://hub.baai.ac.cn/view/21165
9)标题:斯坦福、康奈尔、宾夕法尼亚、华盛顿大学联合 | CREPE: Can Vision-Language Foundation Models Reason Compositionally?(CREPE:视觉语言基础模型能否进行组合推理?)
https://arxiv.org/pdf/2212.07796.pdf
https://hub.baai.ac.cn/view/22629
10)标题:搞多模态不了解最新进展?中科院自动化所撰文首个视觉-语言预训练综述
https://arxiv.org/pdf/2202.09061.pdf
https://hub.baai.ac.cn/view/15943
11)标题:英国牛津大学、萨里大学| Multimodal Learning with Transformers: A Survey(综述:Transformers 多模态学习)
https://arxiv.org/pdf/2206.06488
https://hub.baai.ac.cn/view/18006
12)标题:韩国科学技术院、NAVER公司 | DialogCC: Large-Scale Multi-Modal Dialogue Dataset(DialogCC:大规模多模态对话数据集)
https://arxiv.org/pdf/2212.04119.pdf
https://hub.baai.ac.cn/view/22463
13)标题:斯坦福大学、Meta、华盛顿大学 | Retrieval-Augmented Multimodal Language Modeling(检索增强的多模态语言建模)
https://arxiv.org/pdf/2211.12561.pdf
https://hub.baai.ac.cn/view/21934
14)标题:清华大学、字节跳动 | Multimodal Entity Tagging with Multimodal Knowledge Base(基于多模态知识库的多模态实体标注)
https://arxiv.org/pdf/2201.00693.pdf
https://hub.baai.ac.cn/view/13750
15)标题:谷歌 | End-to-end Generative Pretraining for Multimodal Video Captioning(多模态视频字幕的端到端生成预训练)
https://arxiv.org/pdf/2201.08264
https://hub.baai.ac.cn/view/14288
16)标题:谷歌|LaMDA: Language Models for Dialog Applications(LaMDA:对话应用的语言模型)
https://arxiv.org/pdf/2201.08239.pdf
https://hub.baai.ac.cn/view/14312
17)标题:微软、英伟达|Using DeepSpeed and Megatron to Train Megatron-Turing NLG 530B, A Large-Scale Generative Language Model(使用 DeepSpeed 和 Megatron 训练 Megatron-Turing NLG 5300亿,一种大规模生成语言模型)
18)标题:谷歌 | PaLM: Scaling Language Modeling with Pathways(PaLM:基于Pathways系统扩展语言模型)
https://arxiv.org/pdf/2204.02311v1.pdf
https://hub.baai.ac.cn/view/16161
19)标题:CMU | reStructured Pre-training(重构预训练)
https://arxiv.org/pdf/2206.11147.pdf
https://hub.baai.ac.cn/view/18255
20)标题:Huggingface等 | BLOOM: A 176B-Parameter Open-Access Multilingual Language Model(BLOOM: 176B参数的开源多语言模型)
https://arxiv.org/pdf/2211.05100.pdf
https://hub.baai.ac.cn/view/18992
21)标题:Meta | Atlas: Few-shot Learning with Retrieval Augmented Language Models(Atlas: 用检索增强的语言模型进行小样本学习)
https://arxiv.org/pdf/2208.03299.pdf
https://hub.baai.ac.cn/view/19716
22)标题:谷歌 | Scaling Instruction-Finetuned Language Models(扩展指令微调语言模型)
https://arxiv.org/pdf/2210.11416.pdf
https://hub.baai.ac.cn/view/21163
23)标题:UIUC、谷歌 | Large Language Models Can Self-Improve(大语言模型可以自我改进)
https://arxiv.org/pdf/2210.11610v2.pdf
https://hub.baai.ac.cn/view/21292
24)标题: OpenAI | ChatGPT
https://hub.baai.ac.cn/view/22253
25)标题:以色列理工学院 | Temporal Attention for Language Models(语言模型的时间注意机制)
https://arxiv.org/pdf/2202.02093.pdf
https://hub.baai.ac.cn/view/14618
26)标题:复旦大学、华东师范 | Black-Box Tuning for Language-Model-as-a-Service(语言模型即服务的黑盒调优)
https://arxiv.org/pdf/2201.03514.pdf
https://hub.baai.ac.cn/view/14009
27)标题:清华大学、北京国家信息科学技术研究中心等联合 | Prompt Tuning for Discriminative Pre-trained Language Models(判别式预训练语言模型的提示调优)
https://arxiv.org/pdf/2205.11166
https://hub.baai.ac.cn/view/17415
28)标题:德国拜罗伊特大学 | A deep unsupervised language model for protein design(用于蛋白设计的深度无监督语言模型)
https://doi.org/10.1101/2022.03.09.483666
https://hub.baai.ac.cn/view/15617
29)标题:Meta、UC伯克利、纽约大学 | Learning inverse folding from millions of predicted structures(从数百万个预测蛋白质结构中学习逆折叠,预测序列信息)
https://doi.org/10.1101/2022.04.10.487779
https://hub.baai.ac.cn/view/16410
30)标题:Meta AI | Language models of protein sequences at the scale of evolution enable accurate structure prediction(进化尺度上的蛋白质序列语言模型使准确的结构预测成为可能)
https://doi.org/10.1101/2022.07.20.500902
https://hub.baai.ac.cn/view/19108
31)标题:微软|DeepNet: Scaling Transformers to 1,000 Layers(DeepNet:将Transformer扩展到1000层)
https://arxiv.org/pdf/2203.00555.pdf
https://hub.baai.ac.cn/view/15195
32)标题:谷歌 | Pathways: Asynchronous Distributed Dataflow for ML(Pathways: 用于ML的异步分布式数据流)
https://arxiv.org/pdf/2203.12533.pdf
https://hub.baai.ac.cn/view/15984
33)标题:DeepMind、牛津、IDSIA等|A Generalist Neural Algorithmic Learner(通才神经算法学习者)
https://arxiv.org/pdf/2209.11142v1.pdf
https://hub.baai.ac.cn/view/20678
34)标题:微软 | Foundation Transformers(基础Transformers)
https://arxiv.org/pdf/2210.06423v1.pdf
https://hub.baai.ac.cn/view/21036
35)标题:美国弗吉尼亚大学、微软 | Active Data Pattern Extraction Attacks on Generative Language Models(对生成语言模型的主动数据模式提取攻击)
https://arxiv.org/pdf/2207.10802.pdf
https://hub.baai.ac.cn/view/19332
36)标题:美国石溪大学、IBM | Attention Hijacking in Trojan Transformers(特洛伊木马Transformers 中的注意力劫持)
https://arxiv.org/pdf/2208.04946.pdf
https://hub.baai.ac.cn/view/19640
37)标题:西湖大学、中科院等 | A Survey of Pretraining on Graphs: Taxonomy, Methods, and Applications(图预训练的综述:分类法、方法和应用)
https://arxiv.org/pdf/2202.07893v2
https://hub.baai.ac.cn/view/15946
38)标题:比利时鲁汶大学等 | Measuring Fairness with Biased Rulers: A Survey on Quantifying Biases in Pretrained Language Models(综述:预训练语言模型的量化偏见)
https://arxiv.org/pdf/2112.07447.pdf
https://hub.baai.ac.cn/view/13590
39)标题:加利福尼亚大学 | A Survey on Model Compression for Natural Language Processing(NLP模型压缩综述)
https://arxiv.org/pdf/2202.07105
https://hub.baai.ac.cn/view/14944
40)标题:加利福尼亚大学 | A Survey on Dynamic Neural Networks for Natural Language Processing(NLP动态神经网络综述)
评论
沙发等你来抢