Glitch Tokens in Large Language Models: Categorization Taxonomy and Effective Detection — mentioned by spacecadet

Sun, 26 Jan 2025 02:59:21 GMT

Just used the glitch token attack in a CTF a week ago. The paper is worth a read and there is a repo out there as well that makes the attack straight forward- but implementing it yourself is also something worth doing. I will add that the author thinking no one had done this with deepseek is unlikely, I run this against models every week out of curiosity or for work, not deepseek yet- but considering the adversarial ML community is pretty packed, someone likely had and just didn't write about it. https://arxiv.org/abs/2404.09894 https://arxiv.org/pdf/2410.15052…

Paper and discussion history. A mention is not an endorsement.

Glitch Tokens in Large Language Models: Categorization Taxonomy and Effective Detection · HackerPapers discussions

Glitch Tokens in Large Language Models: Categorization Taxonomy and Effective Detection — mentioned by spacecadet