Recent AI Reading [dd Month yyyy]

Papers

Agentic AI

• AgentStore: Scalable Integration of Heterogeneous Agents As Specialized Generalist Computer Assistant
  • Source code in GitHub
• Teaching Embodied Reinforcement Learning Agents: Informativeness and Diversity of Language Use
  • Teachable_RL source code.

AI Alignment with Human Feedback and Preferences, and other methods

• Reference-Guided Verdict: LLMs-as-Judges in Automatic Evaluation of Free-Form Text
• SelfCodeAlign: Self-Alignment for Code Generation
• MA-RLHF: Reinforcement Learning from Human Feedback with Macro Actions
• RLEF: Grounding Code LLMs in Execution Feedback with Reinforcement Learning

Retrieval-Augmented Generation

• RAGulator: Lightweight Out-of-Context Detectors for Grounded Text Generation
• Long Context RAG Performance of Large Language Models
• RAGViz: Diagnose and Visualize Retrieval-Augmented Generation
• M3DocRAG: Multi-modal Retrieval is What You Need for Multi-page Multi-document Understanding

Synthetic Data

• Fictitious Synthetic Data Can Improve LLM Factuality via Prerequisite Learning
• Hunyuan-Large: An Open-Source MoE Model with 52 Billion Activated Parameters by Tencent

  • Synthetic data is all you need? New large MoE from Tencent was trained on 1.5 trillion tokens of synthetic data. The 389B-A52B MoE outperforms @AIatMeta Llama 3.1 405B across academic benchmarks. [ref]

Miscellaneous

• Large Language Models Reflect the Ideology of their Creators
• Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To!
• Datasets for Large Language Models: A Comprehensive Survey
  • GitHub Repo: Awesome-LLMs-Datasets
• Large Language Models and the Degradation of the Medical Record
• DreamClear: High-Capacity Real-World Image Restoration with Privacy-Safe Dataset Curation
• A Survey of Small Language Models
• Document Parsing Unveiled: Techniques, Challenges, and Prospects for Structured Information Extraction
• A Survey of Small Language Models
• OpenCoder: The Open Cookbook for Top-Tier Code Large Language Models

Articles and Blog Posts

• Creating a LLM-as-a-Judge That Drives Business Results
• Understanding LLMs from Scratch Using Middle School Math

Miscellaneous

• awesome-production-llm - A curated list of awesome open-source libraries for production LLM.
• A New Chapter for fast.ai: How To Solve It With Code