Become a sponsor to LLMs Research
LLMs Research builds open-source implementations of AI research papers. When a paper drops without code, or with code that doesn't quite work, we build it, test the claims, and publish what we find. We also run a newsletter read by over 6000+ practitioners covering what's worth paying attention to in LLM research each week.
Recent and notable projects:
- 🍌 PaperBanana: Open-source implementation of Google Research's agentic framework for generating publication-quality academic diagrams. Includes CLI, Python API, and MCP server for IDE integration.
- 🔍 ReaRAG: Knowledge-guided reasoning with iterative retrieval-augmented generation for improved factual accuracy.
- 📐 SCONE: Implementation and evaluation of the Scaling Embedding Layers paper.
- 🧬 DarwinLM: Evolutionary structured pruning of large language models.
- ⚡ CoupledAdam: Better embeddings via coupled optimizer dynamics.
- 💬 semantic-model-toolkit: Query databases in natural language without writing SQL.
We regularly implement new papers and are also working on original research, including a submission to TMLR.
What sponsorship supports
This is a solo operation. Sponsorship goes directly toward cloud compute for running large models and experiments, access to papers and datasets, and the time spent implementing, testing, and documenting code that would otherwise not exist as open source. Everything we build stays freely available under MIT.
5 sponsors can cover our monthly expenditure on cloud, subscriptions, and other tools we use to implement and publish research papers.
Featured work
-
llmsresearch/paperbanana
Open source implementation and extension of Google Research’s PaperBanana for automated academic figures, diagrams, and research visuals, expanded to new domains like slide generation.
Python 835 -
llmsresearch/rearag
Implementing ReaRAG, a knowledge-guided reasoning model that enhances factual accuracy using iterative retrieval-augmented generation. Adapting the methodology for modular LLM integration and custo…
Python 15 -
llmsresearch/scone
Implementation and evaluation of Scaling Embedding Layers in Language Models research paper
Python 12 -
llmsresearch/darwinlm
Implementation of research paper: DarwinLM: Evolutionary Structured Pruning of Large Language Models
Python 8 -
llmsresearch/coupledadam
Implementation of Better Embeddings with Coupled Adam research paper
Python 7 -
llmsresearch/semantic-model-toolkit
Python library to generate semantic model to use in talk to database application to ask questions in natural language and receive direct answers without writing SQL.
Python 4