arXiv cs.CL
· Papers
Trustworthiness Costs of Domain Adaptation in Small Language Models:A Cross-Architecture Empirical Study
arXiv:2608.00042v1 Announce Type: new Abstract: Domain adaptation of small language models (SLMs) has emerged as a practical strategy for deploying capable NLP systems in resource-constrained, high-stakes environments including healthcare, legal services, and financial analysis. While performance gains from parameter-e