Alternatives and similar repositories for BELLS

Users that are interested in BELLS are comparing it to the libraries listed below

• UKGovernmentBEIS / control-arena
  ControlArena is a collection of settings, model organisms and protocols - for running control experiments.
  ☆90Updated this week
• UKGovernmentBEIS / inspect_evals
  Collection of evals for Inspect AI
  ☆215Updated this week
• UKGovernmentBEIS / inspect_ai
  Inspect: A framework for large language model evaluations
  ☆1,293Updated this week
• METR / task-standard
  METR Task Standard
  ☆159Updated 7 months ago
• TransformerLensOrg / CircuitsVis
  Mechanistic Interpretability Visualizations using React
  ☆282Updated 8 months ago
• ndif-team / nnsight
  The nnsight package enables interpreting and manipulating the internals of deep learned models.
  ☆646Updated this week
• ai-safety-foundation / sparse_autoencoder
  Sparse Autoencoder for Mechanistic Interpretability
  ☆260Updated last year
• centerforaisafety / Intro_to_ML_Safety
  ☆74Updated 2 years ago
• anthropics / sleeper-agents-paper
  Contains random samples referenced in the paper "Sleeper Agents: Training Robustly Deceptive LLMs that Persist Through Safety Training".
  ☆112Updated last year
• ARBORproject / arborproject.github.io
  ☆81Updated 6 months ago
• haizelabs / redteaming-resistance-benchmark
  ☆45Updated last year
• stanfordnlp / pyvene
  Stanford NLP Python library for understanding and improving PyTorch models via interventions
  ☆800Updated this week
• JasonGross / guarantees-based-mechanistic-interpretability
  ☆16Updated this week
• haizelabs / dspy-redteam
  Red-Teaming Language Models with DSPy
  ☆211Updated 6 months ago
• TransluceAI / observatory
  A toolkit for describing model features and intervening on those features to steer behavior.
  ☆198Updated 9 months ago
• GraySwanAI / nanoGCG
  A fast + lightweight implementation of the GCG algorithm in PyTorch
  ☆278Updated 3 months ago
• google-deepmind / dangerous-capability-evaluations
  ☆57Updated last month
• centerforaisafety / HarmBench
  HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
  ☆713Updated last year
• andyrdt / refusal_direction
  Code and results accompanying the paper "Refusal in Language Models Is Mediated by a Single Direction".
  ☆262Updated 2 months ago
• ArthurConmy / Automatic-Circuit-Discovery
  ☆238Updated 11 months ago
• emergent-misalignment / emergent-misalignment
  ☆201Updated 5 months ago
• AlignmentResearch / tuned-lens
  Tools for understanding how transformer predictions are built layer-by-layer
  ☆521Updated 3 weeks ago
• hijohnnylin / neuronpedia
  open source interpretability platform 🧠
  ☆375Updated this week
• UKGovernmentBEIS / hibayes
  ☆31Updated this week
• EleutherAI / sparsify
  Sparsify transformers with SAEs and transcoders
  ☆613Updated last week
• compl-ai / compl-ai
  An open-source compliance-centered evaluation framework for Generative AI models
  ☆161Updated last week
• callummcdougall / sae_vis
  Create feature-centric and prompt-centric visualizations for sparse autoencoders (like those from Anthropic's published research).
  ☆216Updated 8 months ago
• jbloomAus / SAELens
  Training Sparse Autoencoders on Language Models
  ☆935Updated this week
• alan-cooney / transformer-from-scratch
  Decoder only transformer, built from scratch with PyTorch
  ☆31Updated last year
• callummcdougall / ARENA_2.0
  Resources for skilling up in AI alignment research engineering. Covers basics of deep learning, mechanistic interpretability, and RL.
  ☆225Updated 3 weeks ago