2025/12/17 by Rathore, Darshita, Kumar, Vineet, Bansal, Chetna +1
Computer Science · #Artificial Intelligence (cs.AI) #Computation and Language (cs.CL) #Domain (mathematical analysis) #Domain Adaptation and Few-Shot Learning #Domain adaptation #FOS: Computer and information sciences #Machine Learning (cs.LG) #Multimodal Machine Learning Applications #Perspective (graphical) #Precision and recall #Rank (graph theory) #Recall #Robustness (evolution) #Topic Modeling
paper · open access · doi:10.48550/arxiv.2512.15634
published in arXiv (Cornell University) (Cornell University)
openalex publication_date 2025/12/17 · openalex created_date 2025/12/19 · openalex updated_date 2026/07/28
Large language models are increasingly adapted to downstream tasks through fine-tuning. Full supervised fine-tuning (SFT) and parameter-efficient fine-tuning (PEFT) methods, such as Low-Rank Adaptation (LoRA), are two dominant approaches. While PEFT methods are widely used for their computational efficiency, the implications of their configurations (e.g., rank) remain under-explored in downstream Q&A tasks and generalisation. In this work, we perform a comprehensive evaluation across multiple reasoning and recall datasets, conducting a rank sweep to quantify the trade-off between SFT and PEFT. We also compare the accuracy of PEFT and SFT models across in-domain and out-of-domain adaptation, highlighting distinct generalisation behaviour and task-specific forgetting. We demonstrate that LoRA achieves competitive and in some cases superior performance compared to SFT, particularly on reasoning tasks at specific rank values. Additionally, we analyze the internal representations via spectral features and layer-wise attention structures, offering insights into representational drift and structural changes in attention patterns.