3 papers
cs.LG2026
Dialectics of Alignment: Harnessing Unsafe Knowledge for Dynamic Safety Routing
Maryam Hashemzadeh, Jerry Huang, Minseon Kim +2
The prevailing paradigm in large language model (LLM) alignment operates via erasure, filtering unsafe data or training models to strictly refuse harmful prompts. While effective a…
cs.CL2026
Probabilistic Calibration Is a Trainable Capability in Language Models
Davide Baldelli, Sruthi Kuriakose, Maryam Hashemzadeh +2
Language models are increasingly used in settings where outputs must satisfy user-specified randomness constraints, yet their generation probabilities are often poorly calibrated t…
cs.AI2026
REAM: Merging Improves Pruning of Experts in LLMs
Saurav Jha, Maryam Hashemzadeh, Ali Saheb Pasand +3
Mixture-of-Experts (MoE) large language models (LLMs) are among the top-performing architectures. The largest models, often with hundreds of billions of parameters, pose significan…