CLLGFeb 13, 2023

Task-Specific Skill Localization in Fine-tuned Language Models

arXiv:2302.06600v2110 citationsh-index: 82
Originality Incremental advance
AI Analysis

This addresses the need to understand and improve the efficiency and interpretability of fine-tuned models for NLP practitioners, though it is incremental as it builds on parameter-efficient fine-tuning concepts.

The paper tackles the problem of identifying where task-specific skills reside in fine-tuned language models, proposing a method that localizes these skills to a very small subset of parameters (~0.01% of the model) responsible for over 95% of performance, with improvements such as 40-90% error reduction in calibration and better out-of-distribution predictions.

Pre-trained language models can be fine-tuned to solve diverse NLP tasks, including in few-shot settings. Thus fine-tuning allows the model to quickly pick up task-specific ``skills,'' but there has been limited study of where these newly-learnt skills reside inside the massive model. This paper introduces the term skill localization for this problem and proposes a solution. Given the downstream task and a model fine-tuned on that task, a simple optimization is used to identify a very small subset of parameters ($\sim0.01$% of model parameters) responsible for ($>95$%) of the model's performance, in the sense that grafting the fine-tuned values for just this tiny subset onto the pre-trained model gives performance almost as well as the fine-tuned model. While reminiscent of recent works on parameter-efficient fine-tuning, the novel aspects here are that: (i) No further re-training is needed on the subset (unlike, say, with lottery tickets). (ii) Notable improvements are seen over vanilla fine-tuning with respect to calibration of predictions in-distribution ($40$-$90$% error reduction) as well as the quality of predictions out-of-distribution (OOD). In models trained on multiple tasks, a stronger notion of skill localization is observed, where the sparse regions corresponding to different tasks are almost disjoint, and their overlap (when it happens) is a proxy for task similarity. Experiments suggest that localization via grafting can assist certain forms of continual learning.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes