A Comprehensive Survey of Parameter-Efficient Fine-Tuning for Large Language and Vision Models
Bibliographic record
Abstract
The rapid scaling of pre-trained foundation models in natural language processing (NLP), computer vision (CV), and multimodal learning has led to growing interest in methods that can adapt these large models efficiently without incurring the full computational or storage costs of traditional finetuning. Parameter-Efficient Fine-Tuning (PEFT) methods address this challenge by modifying or introducing a small subset of learnable parameters while keeping the majority of the model frozen. In this survey, we present a comprehensive and systematic overview of the landscape of PEFT approaches. We categorize the main families of PEFT methods-including prompt tuning, adapter tuning, low-rank adaptation (e.g., LoRA), BitFit, and sparse updating-providing unified mathematical formulations, detailed comparative analyses, and extensive discussion of their theoretical underpinnings and empirical properties. We also explore implementation considerations, evaluation benchmarks, and real-world applications across language, vision, and multimodal domains. Finally, we highlight open challenges, interpretability gaps, and future research directions in this rapidly evolving field. Our goal is to serve as a foundation for researchers and practitioners seeking to understand, apply, or advance the state of the art in parameterefficient adaptation of large-scale models.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.012 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.002 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.003 |
| Open science | 0.003 | 0.002 |
| Research integrity | 0.002 | 0.003 |
| Insufficient payload (model declined to judge) | 0.006 | 0.004 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".