@inproceedings{liu-etal-2026-look,
    title = "Look Within or Beyond? A Theoretical Comparison Between Parameter-Efficient and Full Fine-Tuning",
    author = "Liu, YongKang  and
      Xu, Xingle  and
      Nie, Ercong  and
      Wang, Zijing  and
      Feng, Shi  and
      Wang, Daling  and
      Li, Qian  and
      Schuetze, Hinrich",
    editor = "Liakata, Maria  and
      Moreira, Viviane P.  and
      Zhang, Jiajun  and
      Jurgens, David",
    booktitle = "Proceedings of the 64th Annual Meeting of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
    month = jul,
    year = "2026",
    address = "San Diego, California, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl/2026.acl-long.2208/",
    pages = "47800--47820",
    ISBN = "979-8-89176-390-6",
    abstract = "\textbf{P}arameter-\textbf{E}fficient \textbf{F}ine-\textbf{T}uning (\textbf{PEFT}) has become a popular alternative to \textbf{F}ull-Parameter \textbf{F}ine-\textbf{T}uning (\textbf{FFT}), achieving similar performance on many benchmarks with far lower computational and memory costs. Yet, its effectiveness on complex tasks such as reasoning and instruction-following remains unclear. In this work, we provide a theoretical and empirical comparison of PEFT and FFT in terms of representational capacity and robustness. We show that PEFT{'}s solution space is a strict subset of FFT{'}s and derive upper bounds revealing how its restricted parameterization limits expressiveness and increases vulnerability to perturbations. Experiments on 20 datasets and 11 adversarial test sets support these findings, indicating that while PEFT performs well on standard tasks, its weaknesses on complex and adversarial settings call for new directions beyond current PEFT paradigms.The source code is in the anonymous GitHub repository[https://anonymous.4open.science/r/PEFTEval-E2AC ]."
}