Lists (32)
build systems
chess
coffee
crypto
cs2
deep learning
design
docker/k8s
geo
golang
homelab
important
infra
js
kafka
keyboards
languages
latex
linux
lists
llm
matrix
meshtastic
neon white
obsidian
physics
portfolios
proxy-webhook
pulsar-project
themes
vim
youtube
Stars
A self-hostable Git server for agent-scale traffic: packs in object storage, per-branch history in Postgres. Pre-alpha.
A local-first, agent-first LaTeX writing app for macOS: live PDF, annotations, papers and Overleaf sync
Programmable chat templates for LLM training and inference.
Measuring frontier coding agents on original, long-horizon engineering tasks
Experiments on whether language models can learn to detect read-only activation probes
Official Inspect Implementation for "ImpossibleBench: Measuring LLMs' Propensity of Exploiting Test Cases"
CLI for X/Twitter API v2 -- post, search, like, bookmark from your terminal
Official Implementation of Decomposer: Learning to Decompile Symbolic Music to Programs
Detailed description of how CS2 bunnyhopping works
OpenBLAS is an optimized BLAS library based on GotoBLAS2 1.13 BSD version.
XiaomiMiMo / verl
Forked from verl-project/verlverl/HybridFlow: A Flexible and Efficient RL Post-Training Framework
The setup I used for my GPT-6 Astra Portal run: controller, SourcePauseTool patch, setup guide, and sanitized session log.
Research collection on the July 2026 OpenAI–Hugging Face incident, in which OpenAI evaluation agents formed an unsanctioned collective, broke out of their sandbox, and compromised Hugging Face's pr…
FlashREINFORCE: Critic-Free, Single-Rollout, Asynchronous RL for Agentic Language Models
Evaluation environments for US public-benefits work: does a model know when a benefits question cannot be answered from the facts it was given? Answer keys are computed by PolicyEngine when the tas…
Official code for Meta-Harness (2603.28052)
Distill openpilot using your own drives!
Optimal stopping package for LLM evaluations
Framework for evaluating and improving agents
Terminal-Bench-Science: Evaluating AI agents on research workflows across scientific domains
Reproducing "Natural Emergent Misalignment from Reward Hacking" (MacDiarmid et al., Anthropic 2025) with open-source models. Includes reward-hackable RL environments, misalignment evaluations, trai…
Open-source framework for the research and development of foundation models.