Stars
Scalable toolkit for efficient model reinforcement
LMCache: Supercharge Your LLM with the Fastest KV Cache Layer
autonomous red teaming platform; multi-agent offensive-security meta-harness
Companion code for the global workspace interpretability paper
Open-source AI-augmented offensive security harness. 13+ autonomous agents, 150+ LLM providers, 5,300+ models, 7,600+ Ed25519-signed attack skills, 56+ built-in tools, 176+ MCP tools. MITRE ATT&CK,…
A minimal LLM-powered zero-day vulnerability scanner by AISLE.
OMCBench is a benchmark suite for evaluating malicious-code detection capabilities. The benchmark consists of a labeled set of 800 Python and JavaScript packages: 400 benign and 400 malicious packa…
Agentic RL on Any Harness at Scale
Бенчмарк для оценки способности LLM-моделей писать код на 1С
DiffusionBlocks: Block-wise Neural Network Training via Diffusion Interpretation
Fast and memory-efficient classical machine learning operators
The batteries-included agent harness.
Reference code for the Meta-Harness paper.
Stable Looped Models and their Scaling Laws
A collection of inspiring lists, manuals, cheatsheets, blogs, hacks, one-liners, cli/web tools and more.
Create real electronics with Typescript and React
GitNexus: The Zero-Server Code Intelligence Engine - GitNexus is a client-side knowledge graph creator that runs entirely in your browser. Drop in a git repository (Github, Gitlab, Azure, Local) or…
Evolutionary generation of efficient GPU kernels
A Python library for extracting structured information from unstructured text using LLMs with precise source grounding and interactive visualization.
D2 is a modern diagram scripting language that turns text to diagrams.
Hera makes Python code easy to orchestrate on Argo Workflows through native Python integrations. It lets you construct and submit your Workflows entirely in Python. ⭐️ Remember to star!