Large Language Model (LLM) is a deep learning-based natural language processing model that learns the syntax and semantics of natural language to generate human-readable text. Large language models are used in many fields, including natural language processing (NLP), computer vision, audio and speech processing, biology, and more, as well as modeling and representation learning, generative, and language translation tasks between multiple media.
The explosion of ChatGPT has made more people pay attention to large models, we have sorted out papers related to large models according to this year's AI summit conference, and the list of top papers on large models is as follows (due to space constraints, this article only shows some top conference papers, click to read the original article to go directly to the top conference list to view all papers)
1.WebGLM: Towards An Efficient Web-Enhanced Question Answering System with Human Preferences
2.Distilling Semantic Concept Embeddings from Contrastively Fine-Tuned Language Models
3.Generative Relevance Feedback with Large Language Models
4.Benchmarking Middle-Trained Language Models for Neural Search
5.What Makes Pre-trained Language Models Better Zero/Few-shot Learners?
6.GLM-130B: An Open Bilingual Pre-trained Model
7.Self-Consistency Improves Chain of Thought Reasoning in Language Models
8.Large Language Models Are Human-Level Prompt Engineers
9.EfficientViT: Memory Efficient Vision Transformer with Cascaded Group Attention
10.Discovering Latent Knowledge in Language Models Without Supervision
11.Language Modelling with Pixels
12.Is Reinforcement Learning (Not) for Natural Language Processing?: Benchmarks, Baselines, and Building Blocks for Natural Language Policy Optimization
13.Learning on Large-scale Text-attributed Graphs via Variational Inference
14.Generate rather than Retrieve: Large Language Models are Strong Context Generators
15.Selection-Inference: Exploiting Large Language Models for Interpretable Logical Reasoning
16.Selective Annotation Makes Language Models Better Few-Shot Learners
17.PEER: A Collaborative Language Model
18.ReAct: Synergizing Reasoning and Acting in Language Models
19.Reward Design with Language Models
20.Automatic Chain of Thought Prompting in Large Language Models
21.UniMax: Fairer and More Effective Language Sampling for Large-Scale Multilingual Pretraining
22.Quantifying Memorization Across Neural Language Models
23.Compositional Semantic Parsing with Large Language Models
24.Ask Me Anything: A simple strategy for prompting language models
25.Language Models are Multilingual Chain-of-Thought Reasoners
26.Learning to Jointly Share and Prune Weights for Grounding Based Vision and Language Models
27.Generating Sequences by Learning to Self-Correct
28.Meta Learning to Bridge Vision and Language Models for Multimodal Few-Shot Learning
29.Visual Classification via Description from Large Language Models
30.Recitation-Augmented Language Models
31.Language Models are Realistic Tabular Data Generators
32.Training language models for deeper understanding improves brain alignment
33.Progressive Prompts: Continual Learning for Language Models without Forgetting
34.Can discrete information extraction prompts generalize across language models?
35.On Pre-training Language Model for Antibody
36.Open-Vocabulary Object Detection upon Frozen Vision and Language Models
37.Mass-Editing Memory in a Transformer
38.Language Models Can Teach Themselves to Program Better
39.Out-of-Distribution Detection and Selective Generation for Conditional Language Models
40.Compositional Task Representations for Large Language Models
41.Planning with Large Language Models for Code Generation
42.Prototypical Calibration for Few-shot Learning of Language Models
43.Multi-lingual Evaluation of Code Generation Models
44.Planning with Language Models through Iterative Energy Minimization
45.Dataless Knowledge Fusion by Merging Weights of Language Models
46.Task Ambiguity in Humans and Language Models
47.Language Models Can (kind of) Reason: A Systematic Formal Analysis of Chain-of-Thought
48.Sub-Task Decomposition Enables Learning in Sequence to Sequence Tasks
49.Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 small
50.Leveraging Large Language Models for Multiple Choice Question Answering
How to use ChatPaper?
In order to allow more researchers to obtain literature knowledge more efficiently, AMiner has developed Chatpaper based on the GLM-130B large model capability to help researchers quickly improve the efficiency of retrieval and reading of papers, obtain the latest research trends in the field, and make scientific research work more comfortable.
ChatPaper is a conversational private knowledge base that integrates retrieval, reading, and knowledge Q&A, and AMiner hopes to use the power of technology to make people more efficient in acquiring knowledge.
ChatPaper：https://www.aminer.cn/chat/g