Showing 2026Show all
2 papers · 1 filter
cs.CL2026
REPRO: Reflective Evolutionary Prompt Optimizer for Automated LLM Red-Teaming
Abrar Alotaibi, Moataz Ahmed
Safety-aligned large language models (LLMs) remain vulnerable to adversarial prompts that bypass their safety training. Existing automated red-teaming methods either require privil…
cs.LG2026
LILA: Calibration-Free Structured Pruning of Large Language Models via Latent Spectral Geometry
Sankar Behera, Dhruv Singh, Anshika Agnihotri +3
Structured pruning of large language models (LLMs) offers hardware-efficient compression, yet existing methods require calibration data, gradient computation, or large auxiliary po…