ArXiv TLDR

Xiao Li

3 papers · Latest:

Cryptography & Security

Usability as a Weapon: Attacking the Safety of LLM-Based Code Generation via Usability Requirements

This paper introduces UPAttack, demonstrating how usability requirements can force LLMs to generate insecure code, achieving up to 98.1% attack success.

2605.10133
Mesoscale & Nanoscale Physics

Tunable high-Chern-number Chern insulators in rhombohedral tetralayer graphene/hBN moiré superlattices

This paper explores tunable high-Chern-number Chern insulators in rhombohedral tetralayer graphene/hBN moiré superlattices, revealing new topological states.

2604.26643
Machine Learning

Nexus: Same Pretraining Loss, Better Downstream Generalization via Common Minima

Nexus optimizer improves LLM generalization by encouraging common minima across data sources, achieving better downstream performance with the same pretraining loss.

2604.09258

📬 Weekly AI Paper Digest

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