ArXiv TLDR

Zheng Lin

4 papers · Latest:

Cryptography & Security

Re-Triggering Safeguards within LLMs for Jailbreak Detection

This paper introduces an embedding disruption method to re-trigger LLM safeguards, effectively detecting and defending against jailbreak attacks.

2605.10611
Cryptography & Security

Guaranteed Jailbreaking Defense via Disrupt-and-Rectify Smoothing

DR-Smoothing offers a guaranteed defense against LLM jailbreaking attacks by disrupting and rectifying prompts, balancing safety and helpfulness.

2605.10582
Machine Learning

Near-Future Policy Optimization

NPO and AutoNPO enhance Reinforcement Learning with Verifiable Rewards (RLVR) by leveraging near-future policy checkpoints for improved off-policy learning.

2604.20733
Machine Learning

Self-Distilled RLVR

RLSD combines RLVR with self-distillation to provide fine-grained updates and reliable directions, improving LLM training stability and convergence.

2604.03128

📬 Weekly AI Paper Digest

Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.