Dev
GitHub repos gaining traction - what high-signal users are starring and what's climbing the board, captured daily and enriched from GitHub. Raw material for spotting new tech and patterns worth building on.
3,195
repos tracked
105
surfaced this week
193
created < 30d
Python
top language
199 repos
-
Prototyping on AI Foundry's ET-SoC-1 (1088-core RISC-V): sys_emu setup, hello world, tensor-unit matmul FLOPs/W vs A100
-
Our library for RL environments + evals
-
RL-fine tuning with GRPO, all in your browser
-
UCCL is an efficient communication library for GPUs, covering collectives, P2P (e.g., KV cache transfer, RL weight transfer), and EP (e.g., GPU-driven)
-
Implement a reasoning LLM in PyTorch from scratch, step by step
-
Train transformer language models with reinforcement learning.
-
GTSAM is a library of C++ classes that implement smoothing and mapping (SAM) in robotics and vision, using factor graphs and Bayes networks as the underlying computing paradigm rather than sparse matrices.
-
Turn any AI agent into an AI Scientist. The #1 Agent Skills library for science, used by 250,000+ scientists worldwide. 177 ready-to-use validated skills plus 100+ scientific databases covering biology, chemistry, medicine, and drug discovery. Compatible with Cursor, Claude Code, Codex, Pi, Antigravity, and the open Agent Skills standard.
-
Autoresearch agents that improve your benchmark: independent re-measurement, enforced budgets, results as pull requests
-
an rlm-style training harness
-
A spy satellite simulator in your browser, except the data is real. Live open source spatial intelligence on a photorealistic 3D globe.
-
Floating macOS sidebar for Microsoft Edge profiles — focus existing windows or open URLs in the chosen profile.
-
Isaac Lab API, powered by MuJoCo-Warp, for RL and robotics research
-
Gym-Anything for Robotics: framework and the RoboWorld benchmark
-
Textbook on reinforcement learning from human feedback
-
Reverse engineer anything with agents, from app behavior down to native binaries.
-
Code for 'Inference-Time Enhancement of Generative Robot Policies via Predictive World Modeling'
-
StarVLA: A Lego-like Codebase for Vision-Language-Action Model Developing
-
Agentic RL Training at Scale
-
Native Wayland screenshot and annotation editor for Omarchy and Hyprland
-
Machine Learning and the Physical World module
-
Multimodal RL training framework for diffusion & omni models
-
We're rebuilding and open-sourcing the world’s software from scratch.
-
Point at any URL/YouTube/Podcast or file. Get the gist. CLI and Chrome Extension.
-
Reproducible navigation benchmarking & residual RL fine-tuning for Wanderland
-
The first exam for code agents in the physical world: frozen robot tasks from many benchmark suites, each answered with a program and verified in a separate sandbox. Runs on Harbor.
-
Official CLI and Python SDK for Prime Intellect - access GPU compute, remote sandboxes, RL environments, and distributed training infrastructure for AI development at scale.
-
Executable RL environment for diagnosing and repairing distributed PyTorch training failures, with sandboxed execution and correctness-verified recovery rewards.
-
A standard API for single-agent reinforcement learning environments, with popular reference environments and related utilities (formerly Gym)
-
Two small, deterministic benchmarks for a first Outerloop run
-
RewardBench: the first evaluation tool for reward models.
-
[CoRL 2026] RPL: Learning Robust Humanoid Perceptive Locomotion on Challenging Terrains
-
A shared, persistent world for people and AIs — the networking, persistence, and multi-participant layer over the eidoverse-video toolkit. Two planes: an append-only world log + lossy presence.
-
Contact-UAN: a proprioception-based residual world model for humanoid control
-
RPent: Agentic Infrastructure for the Physical World
-
★ 47 castorini/onboardingOnboarding guide to Jimmy Lin's research group at the University of Waterloo
-
Open, structured corpus of US primary law: 2M+ sections of state statutory codes, the US Code, and state constitutions, plus the ingestion pipeline that builds it. Data on Hugging Face (CC BY 4.0); scrapers Apache-2.0. Quarterly snapshots.
-
An interface library for RL post training with environments.
-
Framework for evaluating and improving agents
-
Hexapod walker: STS3215 prototype, RL training orchestrator (CoreWeave), MuJoCo sim, robot control
-
An interface library for RL post training with environments.
-
Interactive interface for browsing global, full-resolution satellite imagery
-
Titlebars for Omarchy's floating workspaces: a Hyprland plugin forked from hyprbars
-
★ 0 neubig/harborHarbor is a framework for running agent evaluations and creating and using RL environments.
-
ICMI-014: Initial explorations of how the prefix 'As a Christian' affects GPT-2 processing (TransformerLens, activation patching, zero-ablation, robustness checks)
-
Reinforcement Learning from Christian Feedback: 15-seed GRPO replication on Qwen3.5-27B with three Christian-formulated judge rubrics. Reproduction package for ICMI-018.