GenAI_Papers

This project contains a list of interesting research papers in the field of GenAI.

Topics

1. Overview
2. Goals
3. Scope and Context
4. Research Papers
5. Learning Logs

---

Overview

This repository is dedicated to the aggregation and discussion of groundbreaking research in the field of Generative AI.

Generative AI, or GenAI, refers to the subset of artificial intelligence focused on creating new content, ranging from text and images to code and beyond. The collection of papers included herein spans a variety of topics within GenAI, such as Generative Adversarial Networks (GANs), Variational Autoencoders (VAEs), and Transformer-based models.

This compendium serves as a resource for scholars, practitioners, and enthusiasts seeking to advance the state of the art in AI-driven content generation.

Goals

The primary goals of this repository are:

1. Knowledge Consolidation: To centralize seminal and cutting-edge research papers that define and advance the GenAI field.
2. Community Collaboration: To foster a collaborative environment where ideas and findings can be shared, discussed, and critiqued by the Gen AI research community.
3. Innovation Promotion: To inspire and guide new research initiatives and practical applications of GenAI technologies.
4. Interdisciplinary Integration: To encourage the cross-pollination of ideas from diverse fields such as computer science, cognitive psychology, and digital arts to enrich the GenAI research.

Scope and Context

Scope

The scope of this repository is to encompass a wide array of research within GenAI, including but not limited to:

• Theoretical foundations of generative models
• Technical advancements in algorithm design
• Applications of GenAI in various domains (e.g., art, healthcare, software development)
• Ethical considerations and societal impacts of GenAI

Context

The GenAI field is situated at the intersection of multiple disciplines. It leverages deep learning, statistical modeling, and computational creativity to generate novel outputs that can mimic or even surpass human-level creativity in certain aspects. With the rapid pace of advancement in AI, it is crucial to maintain a clear and organized overview of the progress in this area, which this repository aims to provide.

Research Papers

📝 Note: Not in a particular order.

Classification

Category	Papers	Description
Language Models & General AI	1, 10, 11, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 31, 34, 35, 36, 37, 38, 40, 41, 42, 43, 44, 45, 48, 54, 56, 58, 60, 66, 69, 74, 76, 79, 80, 82, 84, 86, 87, 89, 90, 92, 93, 95, 98, 99, 101, 103, 104	Papers related to language models, their applications, ethical considerations, and improvements in training or functionality.
Vision & Language Integration	3, 4, 29, 30, 33, 64	Focusing on the integration of visual data with language models, including vision transformers and text-to-image personalization.
Attention Mechanisms & Transformers	8, 9, 25, 28, 73	Discussing the theory of attention in deep learning and optimization of transformer models.
Music & Creative AI	5	A unique paper on music generation using AI.
High-Resolution Image Synthesis	6, 7, 63	Discussing high-resolution image synthesis using diffusion models and vision transformers.
Efficiency & Scaling in AI	2, 25, 26, 27, 28, 59, 61, 71, 72, 83, 88, 97	Covering AI efficiency in terms of memory, inference, and scaling.
Environmental Impact of AI	12	A unique paper focusing on the environmental impact of AI systems.
Dialog & Interaction-Focused AI	13, 24, 34, 35, 36, 37, 39, 53, 67, 81, 91	Involving dialogue applications and platforms for interactive language agents.
AI Enhancement & Meta-Learning	27, 31, 32, 37, 46, 47, 49, 55, 57, 62, 65, 68, 70, 75, 78, 96	On improving AI capabilities through self-improvement, preference optimization, and distillation.
Miscellaneous AI Applications	29, 30, 33, 50, 52, 77, 85, 94, 100, 102	Discussing niche AI applications like commonsense norms and visual instruction tuning.

Complete List

1. Enhancing CLIP with GPT-4: Harnessing Visual Descriptions as Prompts
2. EfficientViT: Memory Efficient Vision Transformer with Cascaded Group Attention
3. Key-Locked Rank One Editing for Text-to-Image Personalization
4. ELIXR: Towards a general purpose X-ray artificial intelligence system through alignment of large language models and radiology vision encoders
5. Simple and Controllable Music Generation
6. High-Resolution Image Synthesis with Latent Diffusion Models
7. All are Worth Words: A ViT Backbone for Diffusion Models
8. Attention Is All You Need
9. A Mathematical View of Attention Models in Deep Learning
10. Improving Language Understanding by Generative Pre-Training
11. Large Language Models and the Reverse Turing Test
12. Estimating the Carbon Footprint of Bloom, a 176b Parameter Language Model
13. LaMDA: Language Models for Dialog Applications
14. Gorilla: Large Language Model Connected with Massive APIs
15. Foundation Models for Decision Making Problems, Methods, and Opportunities
16. Continual Pre-training of Language Models
17. How Far Can Camels Go? Exploring the State of Instruction Tuning on Open Resources
18. Alpagasus: Training a Better Alpaca with Fewer Data
19. Ethical and social risks of harm from Language Models
20. Holistic Evaluation of Language Models
21. On the Risk of Misinformation Pollution with Large Language Models
22. The Capacity for Moral Self-Correction in Large Language Models
23. HONEST: Measuring Hurtful Sentence Completion in Language Models
24. ReAct: Synergizing Reasoning and Acting in Language Models
25. Efficiently Scaling Transformer Inference
26. Hungry Hungry Hippos: Towards Language Modeling with State Space Models
27. Promptbreeder: Self-Referential Self-Improvement via Prompt Evolution
28. Efficient Streaming Language Models with Attention Sinks
29. Visual Instruction Tuning
30. Improved Baselines with Visual Instruction Tuning
31. Direct Preference Optimization: Your Language Model is Secretly a Reward Model
32. Distil-Whisper: Robust Knowledge Distilation via Large-Scale Pseudo Labelling
33. Reading Books is Great, But Not if You Are Driving! Visually Grounded Reasoning about Defeasible Commonsense Norms
34. TeacherLM: Teaching to Fish Rather Than Giving the Fish, Language Modeling Likewise
35. Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflection
36. InstructRetro: Instruction Tuning post Retrieval-Augmented Pretraining
37. OpenAgents: An Open Platform for Language Agents in the Wild
38. Large Language Models Understand and Can be Enhanced by Emotional Stimuli
39. Communicative Agents for Software Development
40. Large Language Models Are Human-Level Prompt Engineers
41. Least-to-Most Prompting Enables Complex Reasoning in Large Language Models
42. Self-Consistency Improves Chain of Thought Reasoning in Language Models
43. Language Models can be Logical Solvers
44. Lost in the Middle: How Language Models Use Long Contexts
45. Contrastive Chain-of-Thought Prompting
46. RankZephyr: Effective and Robust Zero-Shot Listwise Reranking is a Breeze!
47. LLM in a flash: Efficient Large Language Model Inference with Limited Memory
48. PowerInfer: Fast Large Language Model Serving with a Consumer-grade GPU
49. Human Centered Loss Functions (HALOs)
50. A Challenger to GPT-4V? Early Explorations of Gemini in Visual Expertise
51. Distributed Inference and Fine-tuning of Large Language Models Over The Internet
52. GAIA: Zero-shot Talking Avatar Generation
53. SLEEPER AGENTS: TRAINING DECEPTIVE LLMS THAT PERSIST THROUGH SAFETY TRAINING
54. LLMLingua: Compressing Prompts for Accelerated Inference of Large Language Models
55. Foundations of Vector Retrieval
56. Self-Rewarding Language Models
57. BloombergGPT: A Large Language Model for Finance
58. Mistral 7B
59. Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer
60. MegaBlocks: Efficient Sparse Training with Mixture-of-Experts
61. Mixture-of-Experts Meets Instruction Tuning: A Winning Combination for Large Language Models
62. Orca 2: Teaching Small Language Models How to Reason
63. ConvNets Match Vision Transformers at Scale
64. Emu Video: Factorizing Text-to-Video Generation by Explicit Image Conditioning
65. Pythia: A Suite for Analyzing Large Language Models Across Training and Scaling
66. Llama 2: Open Foundation and Fine-Tuned Chat Models
67. QLoRA: Efficient Finetuning of Quantized LLMs
68. RAG vs Fine-tuning: Pipelines, Tradeoffs, and a Case Study on Agriculture
69. Training language models to follow instructions with human feedback
70. Spotting LLMs With Binoculars: Zero-Shot Detection of Machine-Generated Text
71. Sparse Networks from Scratch: Faster Training without Losing Performance
72. ZeRO: Memory Optimizations Toward Training Trillion Parameter Models
73. Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity
74. DeepSeekMoE: Towards Ultimate Expert Specialization in Mixture-of-Experts Language Models
75. MoE-Mamba: Efficient Selective State Space Models with Mixture of Experts
76. Code Llama: Open Foundation Models for Code
77. LLaMA Pro: Progressive LLaMA with Block Expansion
78. Self-Play Fine-Tuning Converts Weak Language Models to Strong Language Models
79. Monarch Mixer: A Simple Sub-Quadratic GEMM-Based Architecture
80. Large Language Model based Multi-Agents: A Survey of Progress and Challenges
81. Retrieval-Augmented Generation for Large Language Models: A Survey
82. ReAugKD: Retrieval-Augmented Knowledge Distillation For Pre-trained Language Models
83. The Era of 1-bit LLMs: All Large Language Models are in 1.58 Bits
84. Assisting in Writing Wikipedia-like Articles From Scratch with Large Language Models
85. RAPTOR: Recursive Abstractive Processing for Tree-Organized Retrieval
86. Beyond A*: Better Planning with Transformers via Search Dynamics Bootstrapping
87. Datasets for Large Language Models: A Comprehensive Survey
88. An LLM Compiler for Parallel Function Calling
89. Chain-of-Thought Prompting Elicits Reasoning in Large Language Models
90. Plan-and-Solve Prompting: Improving Zero-Shot Chain-of-Thought Reasoning by Large Language Models
91. ReWOO: Decoupling Reasoning from Observations for Efficient Augmented Language Models
92. StructLM: Towards Building Generalist Models for Structured Knowledge Grounding
93. A Critical Evaluation of AI Feedback for Aligning Large Language Models
94. Are More LLM Calls All You Need? Towards Scaling Laws of Compound Inference Systems
95. Are Emergent Abilities of Large Language Models a Mirage?
96. Yi: Open Foundation Models by 01.AI
97. ORPO: Monolithic Preference Optimization without Reference Model
98. Do Large Language Models Understand Logic or Just Mimick Context?
99. Evaluating Large Language Models Trained on Code
100. Self-Refine: Iterative Refinement with Self-Feedback
101. Reflexion: Language Agents with Verbal Reinforcement Learning
102. MM-REACT: Prompting ChatGPT for Multimodal Reasoning and Action
103. HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging Face
104. AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation
105. Megalodon: Efficient LLM Pretraining and Inference with Unlimited Context Length
106. A survey of Generative AI Applications
107. MAC-SQL: A Multi-Agent Collaborative Framework for Text-to-SQL
108. Prometheus 2: An Open Source Language Model Specialized in Evaluating Other Language Models
109. MetaGPT: Meta Programming For A Multi-Agent Collaborative Framework
110. Understanding Transformer Reasoning Capabilities via Graph Algorithms
111. Banishing LLM Hallucinations Requires Rethinking Generalization
112. Your Context Is Not an Array: Unveiling Random Access Limitations in Transformers
113. LLM Lies: Hallucinations are not Bugs, but Features as Adversarial Examples
114. Memory^3 : Language Modeling with Explicit Memory
115. NeuroLogic Decoding: (Un)supervised Neural Text Generation with Predicate Logic Constraints
116. LOTUS: Enabling Semantic Queries with LLMs Over Tables of Unstructured and Structured Data
117. Text2SQL is Not Enough: Unifying AI and Databases with TAG
118. Chain-of-Thought Reasoning Without Prompting
119. Chain of Thought Empowers Transformers to Solve Inherently Serial Problems
120. Premise Order Matters in Reasoning with Large Language Models
121. Teaching Large Language Models to Self-Debug
122. SciAgents: Automating scientific discovery through multi-agent intelligent graph reasoning
123. Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context
124. Agentic Retrieval-Augmented Generation for Time Series Analysis
125. Reflection-Tuning: Data Recycling Improves LLM Instruction-Tuning
126. OLMoE: Open Mixture-of-Experts Language Models
127. Quiet-STaR: Language Models Can Teach Themselves to Think Before Speaking
128. Agent Q: Advanced Reasoning and Learning for Autonomous AI Agents
129. Let's Verify Step by Step
130. Learn Beyond The Answer: Training Language Models with Reflection for Mathematical Reasoning
131. V-STaR: Training Verifiers for Self-Taught Reasoners
132. Agent Workflow Memory
133. Conversational Text-To-SQL: An Odyssey Into State-of-the-art And Challenges Ahead
134. Guiding LLMs The Right Way: Fast, Non-Invasive Constrained Generation
135. Enhancing Few-shot Text-to-SQL Capabilities of Large Language Models: A Study on Prompt Design Strategies
136. Efficient schema-less text-to-SQL conversion using large language models
137. SQL-to-Schema Enhances Schema Linking in Text-to-SQL
138. CodeS: Towards Building Open-source Language Models for Text-to-SQL
139. A Theoretical Understanding of Chain-of-Thought: Coherent Reasoning and Error-Aware Demonstration
140. Principled Instructions Are All You Need for Questioning LLaMA-1/2, GPT-3.5/4
141. Steering Large Language Models between Code Execution and Textual Reasoning
142. Paired Open-Ended Trailblazer (POET): Endlessly Generating Increasingly Complex and Diverse Learning Environments and Their Solutions

Learning Logs

Date	Learning