Pinned Loading

1. DynamicSAEGuardrails DynamicSAEGuardrails Public

   Code for the paper "SAEs Can Improve Unlearning: Dynamic Sparse Autoencoder Guardrails for Precision Unlearning in LLMs"

   Python 8

2. neulab/gemini-benchmark neulab/gemini-benchmark Public

   Jupyter Notebook 151 21

3. transformer-gan transformer-gan Public

   Forked from amazon-science/transformer-gan

   Python

4. poison-set-selection poison-set-selection Public

   Code for Pick Your Poison: Learning to Select Poison Sets for Stronger LLM Backdoor Attacks.

   Python 2

5. defending-against-abliteration defending-against-abliteration Public

   Post-hoc, fine-tuning-free defenses against LLM abliteration / refusal feature ablation.

   Python 1

6. mole mole Public

   Benchmark for detecting insider threats by AI agents in a simulated frontier AI lab

   Python 1