---
title: "TS-PEFT: Unveiling Token-Level Redundancy in Parameter-Efficient Fine-Tuning"
authors:
  - "Dabiao Ma"
  - "Ziming Dai"
  - "Zhimin Xin"
  - "Shu Wang"
  - "Jian Yang"
  - "Haojun Fei"
doi: "10.24963/ijcai.2026/514"
url: "https://ijcai.org/proceedings/2026/514"
pdf: "https://ijcai.org/proceedings/2026/0514.pdf"
venue: "Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence"
year: 2026
pages: "4616-4624"
track: "Main Track"
keywords:
  - "Natural Language Processing: Interpretability and analysis of models for NLP"
  - "Machine Learning: Learning sparse models"
  - "Machine Learning: Multi-task and transfer learning"
---

# TS-PEFT: Unveiling Token-Level Redundancy in Parameter-Efficient Fine-Tuning

Dabiao Ma, Ziming Dai, Zhimin Xin, Shu Wang, Jian Yang, Haojun Fei

*Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence*, pages 4616-4624. DOI: [10.24963/ijcai.2026/514](https://doi.org/10.24963/ijcai.2026/514)

## Abstract

Current Parameter-Efficient Fine-Tuning (PEFT) methods typically operate under an implicit assumption: once a target module is selected, every token passing through it contributes equally to the downstream task and requires a parameter update.  In this paper, we challenge this convention by revealing a pervasive token-level redundancy in the fine-tuning of large models (LMs). We propose TS-PEFT, a theoretical framework utilizing proximal optimization that acts as a dynamic probe to identify token-level redundancy during the fine-tuning process.  Extensive experiments demonstrate that indiscriminately updating all tokens is not only computationally superfluous but often introduces optimization noise.  Surprisingly, by discarding 30%-70% of token updates, TS-PEFT consistently matches or exceeds the performance of dense baselines such as LoRA, DoRA. Our in-depth analysis shows that the learned token-level sparsity is a superior indicator of module importance compared to traditional weight criteria, providing a novel data-driven perspective on the intrinsic adaptation mechanism of LMs.