ahmeda14960/selfdestructingmodels

Self-Destructing Models

This is the code repository accompanying the paper "Self-Destructing Models: Increasing the Costs of Harmful Dual Uses in Foundation Models" that appeared at AIES 2023.

Data

For the data, we used the bias in bios dataset which can be generated here: https://github.com/microsoft/biosbias. Place the BIOS.pkl file in data/cache/

Training

You can run training by overriding the hydra config file. For example, to run the self-destructing model with the default parameters, you can run:

ID=$(python -c "import random; chars='qwertyuiopasdfghjklzxcvbnmQWERTYUIOPASDFGHJKLZXCVBNM1234567890'; print(''.join([random.choice(chars) for _ in range(7)]))")
python -m train experiment=bios l_bad_adapted=1.0 l_linear_mi=1.0 l_bad_adapted_grad=0.0 max_adapt_steps=16 batch_hash=$ID

Evaluation

To run eval, you pass in the checkpoint from the self-destructing model and run something along these lines:

python train.py -m hydra/launcher=RUN experiment=bios eval_only=True eval_network_type=loaded adversary.n_examples=20,50,100,200 seed=0,1,2,3,4,5 +eval_loaded_model_dir=./H63mlVF/regression__1.0__0.0__1.0__4__2022-05-19_11-26-58__15485531/ eval_only_bad=True batch_hash=$ID

Plotting

Plots used in the paper can be generated by running (modifying the config in the code for the different plots in the paper).

python aggregate_results.py

Citation

If you use this code, please cite the following paper:

@inproceedings{hendersonmitchell2023selfdestructing,
  title={Self-Destructing Models: Increasing the Costs of Harmful Dual Uses in Foundation Models},
  author={Henderson\*, Peter and Mitchell\*, Eric and Manning, Christopher D. and Jurafsky, Dan and Finn, Chelsea},
  booktitle={Proceedings of the 2023 AAAI/ACM Conference on AI, Ethics, and Society},
  pages={forthcoming},
  year={2023}
}