Lists (1)
Sort Name ascending (A-Z)
Starred repositories
Agent skill that removes signs of AI-generated writing from text
A framework for state-of-the-art pre-trained bio foundation models on genomics and transcriptomics modalities.
Anonymous code submission for PROMoral.
🦍 Memorilla: Latent Semantic Memory for LLMs
Starter code for running CRUX-type experiments
PostHog context and skill assembly for AI agents and LLM tasks
Build an agent harness and control it end-to-end. Open-source SDK for production AI agents in Python & TypeScript - any model, any cloud.
Object-storage native graph database with branching and typed ontology
Soft-MTP: Multi-Token Prediction Done Right. KL gradient recovery for faster speculative decoding and better language model training.
Mixed-precision numerics benchmarks in Rust and Python - covering GEMMs, SYRKs, DOTs, and higher-level BLAS and LAPACK-style functionality
Mixed-precision math, linear-algebra & retrieval kernels, 4000+ symbols from 4-bit numbers to 128-bit complex, for x86, Arm, RISC-V, LoongArch, Power & WASM, AMX & SME included 📐
Official Pytorch Implementation of MedVTok: A General-Purpose Medical Visual Tokenizer
Paged Attention, Prefix Caching, Continuous Batching and Serving in one file
The harness of Open Math OS, infrastructure for Math 2.0. Built for everyday math research.
The Actor-Gatherer framework: a context management framework for AI agents.
FlyDSL is the Python front‑end of the project: a Flexible Layout Python DSL for expressing tiling, partitioning, data movement, and kernel structure at a high level.
🧩 FrontierAgent, our agent framework, open-sourced alongside it — native command-line TUI, ReAct and Agent Team modes, one command on macOS and Linux, no preinstall, no hard Docker dependency.
Bringing Multi-Head Latent Attention to pretrained LLMs and VLMs: MHA2MLA (ACL 2025), MHA2MLA-VLM (AAAI 2026), and MLAfication.
Bridge Megatron-Core to Hugging Face/Reinforcement Learning
Turn any repository into verifiable RL environments for coding agents - Harbor tasks you can train on, evaluate and share on the Hugging Face Hub
Container recipes and build scripts for LUMI AI Factory container images.