Skip to content

@AlignmentResearch FAR.AI

Change the repository type filter

All

    Repositories list

    61 repositories

    • obfuscation-atlas

      Public
      The Obfuscation Atlas: Mapping Where Honesty Emerges in RLVR with Deception Probes
      rlvrreward-hackingobfuscated-activations
      rlvrreward-hackingobfuscated-activationsobfuscated-policyobfuscation-atlasmbpp-honeypot
      Python
      •2•13•0•5•Updated Jul 11, 2026Jul 11, 2026

    • aim

      Public
      Aim 💫 — An easy-to-use & supercharged open-source experiment tracker.
      Python
      •
      Apache License 2.0
      •400•0•0•5•Updated Jul 10, 2026Jul 10, 2026

    • mlflow

      Public
      Open source platform for the machine learning lifecycle
      Python
      •
      Apache License 2.0
      •6k•0•0•6•Updated Jul 10, 2026Jul 10, 2026

    • far-project-template

      Public
      Template for new experiment repositories at FAR
      Python
      •0•0•0•11•Updated Jul 8, 2026Jul 8, 2026

    • pytorch

      Public
      Tensors and Dynamic neural networks in Python with strong GPU acceleration
      Python
      •
      Other
      •28k•0•0•4•Updated Jul 8, 2026Jul 8, 2026

    • lp_sae

      Public
      Training Sparse Autoencoders on DRC networks
      HTML
      •
      MIT License
      •249•0•0•3•Updated Jul 8, 2026Jul 8, 2026

    • KataGoVisualizer

      Public
      HTML
      •
      MIT License
      •1•3•6•5•Updated Jul 8, 2026Jul 8, 2026

    • defense-in-depth-demo

      Public
      Python
      •
      MIT License
      •1•4•0•4•Updated Jul 8, 2026Jul 8, 2026

    • deception-evasion-honesty

      Public
      Python
      •3•2•1•2•Updated Jul 8, 2026Jul 8, 2026

    • peft

      Public
      🤗 PEFT: State-of-the-art Parameter-Efficient Fine-Tuning.
      Python
      •
      Apache License 2.0
      •2.4k•0•0•2•Updated Jul 8, 2026Jul 8, 2026

    • entangled.py

      Public
      Python port of Entangled
      Python
      •
      Apache License 2.0
      •13•0•0•9•Updated Jul 8, 2026Jul 8, 2026

    • voltage-park-sdk

      Public
      Python
      •
      BSD 3-Clause "New" or "Revised" License
      •0•0•0•5•Updated Jul 8, 2026Jul 8, 2026

    • HarmBench

      Public
      Fork of HarmBench for getting R2D2 working
      Jupyter Notebook
      •
      MIT License
      •147•0•0•4•Updated Jul 8, 2026Jul 8, 2026

    • gcg

      Public
      Fork of official GCG repo
      Python
      •
      MIT License
      •631•0•0•1•Updated Jul 8, 2026Jul 8, 2026

    • scaling-llm-robustness-paper

      Public
      Code used for the paper `Scaling Trends in Language Model Robustness`
      Python
      •
      MIT License
      •0•2•0•5•Updated Jul 8, 2026Jul 8, 2026

    • kueue

      Public
      Kubernetes-native Job Queueing
      Go
      •
      Apache License 2.0
      •687•0•0•3•Updated Jul 8, 2026Jul 8, 2026

    • s3bucket_exporter

      Public
      S3 bucket exporter exposing metrics size and number of objects
      Go
      •
      MIT License
      •15•0•0•2•Updated Jul 8, 2026Jul 8, 2026

    • leela-zero

      Public
      Go engine with no human-provided knowledge, modeled after the AlphaGo Zero paper.
      C++
      •
      GNU General Public License v3.0
      •1k•0•0•5•Updated Jul 8, 2026Jul 8, 2026

    • sae-k-sparse-mamba

      Public
      K-Sparse Autoencoders for Mamba
      Python
      •
      MIT License
      •101•2•0•2•Updated Jul 8, 2026Jul 8, 2026

    • polygames

      Public
      C++
      •
      MIT License
      •55•1•8•5•Updated Jul 8, 2026Jul 8, 2026

    • azure-storage-fuse

      Public
      A virtual file system adapter for Azure Blob storage
      Go
      •
      Other
      •246•0•0•1•Updated Jul 8, 2026Jul 8, 2026

    • kubespray

      Public
      Deploy a Production Ready Kubernetes Cluster
      Jinja
      •
      Apache License 2.0
      •6.9k•0•0•2•Updated Jul 8, 2026Jul 8, 2026

    • posterly-far

      Public
      Build academic conference posters as a single HTML/CSS file, rendered to print-ready PDF via headless Chromium. A coding-agent skill.
      Python
      •
      GNU Affero General Public License v3.0
      •8•0•0•0•Updated Jun 19, 2026Jun 19, 2026

    • tool_robust_exploration

      Public
      A small study exploring LLM instruction hierarchy and techniques for improving prompt robustness
      Python
      •0•0•0•0•Updated Jun 4, 2026Jun 4, 2026

    • impossiblebench

      Public
      Official Inspect Implementation for "ImpossibleBench: Measuring LLMs' Propensity of Exploiting Test Cases"
      Python
      •
      MIT License
      •11•1•0•0•Updated Mar 15, 2026Mar 15, 2026

    • persona_vectors

      Public
      Persona Vectors: Monitoring and Controlling Character Traits in Language Models
      Python
      •109•3•0•4•Updated Feb 20, 2026Feb 20, 2026

    • AttemptPersuadeEval

      Public
      Python
      •
      Apache License 2.0
      •3•13•3•1•Updated Jan 30, 2026Jan 30, 2026

    • accelerate

      Public
      🚀 A simple way to launch, train, and use PyTorch models on almost any device and distributed configuration, automatic mixed precision (including fp8), and easy-…
      Python
      •
      Apache License 2.0
      •1.4k•1•0•0•Updated Nov 25, 2025Nov 25, 2025

    • train-learned-planner

      Public
      Experimenting with CleanRL for learned-planners
      Python
      •
      Other
      •3•9•1•1•Updated Oct 13, 2025Oct 13, 2025

    • cupbearer

      Public
      A library for mechanistic anomaly detection
      Python
      •
      MIT License
      •12•0•0•0•Updated Aug 19, 2025Aug 19, 2025
    ProTip! When viewing an organization's repositories, you can use the props. filter to filter by custom property.