5 papers · 1 filter
SafeMERGE: Preserving Safety Alignment in Fine-Tuned Large Language Models via Selective Layer-Wise Model Merging
Aladin Djuhera, Swanand Ravindra Kadhe, Farhan Ahmed +2
Fine-tuning large language models (LLMs) is a common practice to adapt generalist models to specialized domains. However, recent studies show that fine-tuning can erode safety alig…
STaD: Scaffolded Task Design for Identifying Compositional Skill Gaps in LLMs
Sungeun An, Swanand Ravindra Kadhe, Shailja Thakur +2
Benchmarks are often used as a standard to understand LLM capabilities in different domains. However, aggregate benchmark scores provide limited insight into compositional skill ga…
When Data is the Algorithm: A Systematic Study and Curation of Preference Optimization Datasets
Aladin Djuhera, Farhan Ahmed, Swanand Ravindra Kadhe +3
Aligning large language models (LLMs) is a central objective of post-training, often achieved through reward modeling and reinforcement learning methods. Among these, direct prefer…
Fixing It in Post: A Comparative Study of LLM Post-Training Data Quality and Model Performance
Aladin Djuhera, Swanand Ravindra Kadhe, Syed Zawad +3
Recent work on large language models (LLMs) has increasingly focused on post-training and alignment with datasets curated to enhance instruction following, world knowledge, and spe…
GneissWeb: Preparing High Quality Data for LLMs at Scale
Hajar Emami Gohari, Swanand Ravindra Kadhe, Syed Yousaf Shah +29
Data quantity and quality play a vital role in determining the performance of Large Language Models (LLMs). High-quality data, in particular, can significantly boost the LLM's abil…