Alternatives and similar repositories for sleeper-agents

Users that are interested in sleeper-agents are comparing it to the libraries listed below

• ckkissane / crosscoder-model-diff-replication
  Open source replication of Anthropic's Crosscoders for Model Diffing
  ☆59Updated 11 months ago
• slavachalnev / SAE-TS
  Improving Steering Vectors by Targeting Sparse Autoencoder Features
  ☆24Updated 10 months ago
• EleutherAI / elk-generalization
  Investigating the generalization behavior of LM probes trained to predict truth labels: (1) from one annotator to another, and (2) from e…
  ☆28Updated last year
• UFO-101 / auto-circuit
  A library for efficient patching and automatic circuit discovery.
  ☆77Updated 2 months ago
• ApolloResearch / e2e_sae
  Sparse Autoencoder Training Library
  ☆54Updated 5 months ago
• ApolloResearch / apd
  Attribution-based Parameter Decomposition
  ☆31Updated 3 months ago
• METR / RE-Bench
  ☆102Updated 5 months ago
• steering-vectors / steering-vectors
  Steering vectors for transformer language models in Pytorch / Huggingface
  ☆125Updated 7 months ago
• callummcdougall / sae_visualizer
  ☆29Updated last year
• EleutherAI / delphi
  Delphi was the home of a temple to Phoebus Apollo, which famously had the inscription, 'Know Thyself.' This library lets language models …
  ☆215Updated this week
• ArthurConmy / Automatic-Circuit-Discovery
  ☆242Updated last year
• safety-research / safety-tooling
  Inference API for many LLMs and other useful tools for empirical research
  ☆74Updated last week
• TransformerLensOrg / CircuitsVis
  Mechanistic Interpretability Visualizations using React
  ☆291Updated 9 months ago
• nrimsky / LM-exp
  LLM experiments done during SERI MATS - focusing on activation steering / interpreting activation spaces
  ☆97Updated 2 years ago
• callummcdougall / sae_vis
  Create feature-centric and prompt-centric visualizations for sparse autoencoders (like those from Anthropic's published research).
  ☆221Updated 9 months ago
• neelnanda-io / Crosscoders
  ☆54Updated 10 months ago
• EleutherAI / steering-llama3
  ☆30Updated last year
• UKGovernmentBEIS / control-arena
  ControlArena is a collection of settings, model organisms and protocols - for running control experiments.
  ☆99Updated this week
• redwoodresearch / Easy-Transformer
  ☆124Updated last year
• nrimsky / CAA
  Steering Llama 2 with Contrastive Activation Addition
  ☆187Updated last year
• jbloomAus / SAEDashboard
  ☆71Updated last month
• METR / task-standard
  METR Task Standard
  ☆162Updated 8 months ago
• redwoodresearch / interp
  Redwood Research's transformer interpretability tools
  ☆14Updated 3 years ago
• FlyingPumba / InterpBench
  A benchmark for mechanistic discovery of circuits in Transformers
  ☆16Updated 9 months ago
• amack315 / unsupervised-steering-vectors
  ☆34Updated last year
• annahdo / implementing_activation_steering
  A collection of different ways to implement accessing and modifying internal model activations for LLMs
  ☆19Updated 11 months ago
• adamkarvonen / SAEBench
  ☆125Updated 2 weeks ago
• neelnanda-io / 1L-Sparse-Autoencoder
  ☆127Updated last year
• LoryPack / LLM-LieDetector
  Code for the ICLR 2024 paper "How to catch an AI liar: Lie detection in black-box LLMs by asking unrelated questions"
  ☆72Updated last year
• redwoodresearch / alignment_faking_public
  ☆78Updated 4 months ago