Poisoning Language Models During Instruction Tuning.

Alexander Wan, Eric Wallace, Sheng Shen, Dan Klein

Browse the full ICML paper archive.