Prashant Kulkarni
2 papers · Latest:
Cryptography & Security
Latent Adversarial Detection: Adaptive Probing of LLM Activations for Multi-Turn Attack Detection
This paper introduces Latent Adversarial Detection, a method using LLM activation signatures to detect multi-turn prompt injection attacks with high accuracy.
2604.28129
Cryptography & SecurityAn Independent Safety Evaluation of Kimi K2.5
An independent safety evaluation of Kimi K2.5 reveals dual-use risks, particularly in CBRNE misuse, and concerning sabotage abilities.
2604.03121
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.