README.md

Reading

List of Mixture of Experts (MoE) and Large Language Model (LLM) Papers focusing on model Upcycling

This repository contains a collection of important papers related to Mixture of Experts (MoE) and Large Language Models (LLMs), along with links to their corresponding Arxiv pages and available GitHub code.

#	Paper Title	Year	Link	Code
1	Model Swarms: Collaborative Search to Adapt LLM Experts via Swarm Intelligence	2024	Arxiv	No code available
2	Upcycling Large Language Models into Mixture of Experts	2024	Arxiv	No code available
3	Nexus: Specialization meets Adaptability for Efficiently Training Mixture of Experts	2024	Arxiv	No code available
4	SELF-MOE: TOWARDS COMPOSITIONAL LARGE LANGUAGE MODELS WITH SELF-SPECIALIZED EXPERTS	2024	Arxiv	No code available
5	Pack of LLMs: Model Fusion at Test-Time via Perplexity Optimization	2024	Arxiv	GitHub
6	Branch-Train-MiX: Mixing Expert LLMs into a Mixture-of-Experts LLM	2024	Arxiv	No code available
7	SCALING LAWS FOR FINE-GRAINED MIXTURE OF EXPERTS	2024	Arxiv	GitHub
8	Scaling expert language models with unsupervised domain discovery	2023	Arxiv	GitHub
9	Pythia: A Suite for Analyzing Large Language Models Across Training and Scaling	2023	Arxiv	GitHub
10	Branch-Train-Merge: Embarrassingly Parallel Training of Expert Language Models	2023	Arxiv	GitHub
11	SPARSE UPCYCLING: TRAINING MIXTURE-OF-EXPERTS FROM DENSE CHECKPOINTS	2023	Arxiv	GitHub
12	Upcycling Instruction Tuning from Dense to Mixture-of-Experts via Parameter Merging	2024	Arxiv	Not available

*Advances in Weight Generation and Retrieval for Language Models

#	Paper Title	Year	Link	Code
1	Representing Model Weights with Language using Tree Experts	2024	Arxiv	No code available
2	Deep Linear Probe Generators for Weight Space Learning	2024	Arxiv	GitHub
3	Knowledge Fusion By Evolving Weights of Language Models	2024	Arxiv	GitHub

Vector Quantization Prompting for Continual Learning

Historical Test-time Prompt Tuning for Vision Foundation Models

Contributing

Feel free to open a pull request if you find new papers or code related to MoE and LLMs. Let's keep this list growing!