Bayesian Principles Improve Prompt Learning In Vision-Language Models

Mingyu Kim, Jongwoo Ko, Mijung Park
Proceedings of The 28th International Conference on Artificial Intelligence and Statistics, PMLR 258:4078-4086, 2025.

Abstract

Prompt learning is a popular fine-tuning method for vision-language models due to its efficiency. It requires a small number of additional learnable parameters while significantly enhancing performance on target tasks. However, most existing methods suffer from overfitting to fine-tuning data, yielding poor generalizability. To address this, we propose a new training objective function based on a Bayesian learning principle to balance adaptability and generalizability. We derive a prior over the logits, where the mean function is parameterized by the pre-trained model, while the posterior corresponds to the fine-tuned model. This objective establishes a balance by allowing the fine-tuned model to adapt to downstream tasks while remaining close to the pre-trained model. To avoid the overfitting issues of the standard softmax function, we adopt the one-vs-each softmax approximation along with its Pólya-Gamma augmentation (OVE-PG). We evaluate our method on several benchmark datasets and demonstrate that using the Bayesian principle for prompt learning is indeed a sensible choice. Code is available at the \url{https://github.com/ParkLabML/Bayesian_Principles_Improve_Prompt_Learning_In_Vision_Language_Models.}

Cite this Paper

BibTeX
@InProceedings{pmlr-v258-kim25c, title = {Bayesian Principles Improve Prompt Learning In Vision-Language Models}, author = {Kim, Mingyu and Ko, Jongwoo and Park, Mijung}, booktitle = {Proceedings of The 28th International Conference on Artificial Intelligence and Statistics}, pages = {4078--4086}, year = {2025}, editor = {Li, Yingzhen and Mandt, Stephan and Agrawal, Shipra and Khan, Emtiyaz}, volume = {258}, series = {Proceedings of Machine Learning Research}, month = {03--05 May}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v258/main/assets/kim25c/kim25c.pdf}, url = {https://proceedings.mlr.press/v258/kim25c.html}, abstract = {Prompt learning is a popular fine-tuning method for vision-language models due to its efficiency. It requires a small number of additional learnable parameters while significantly enhancing performance on target tasks. However, most existing methods suffer from overfitting to fine-tuning data, yielding poor generalizability. To address this, we propose a new training objective function based on a Bayesian learning principle to balance adaptability and generalizability. We derive a prior over the logits, where the mean function is parameterized by the pre-trained model, while the posterior corresponds to the fine-tuned model. This objective establishes a balance by allowing the fine-tuned model to adapt to downstream tasks while remaining close to the pre-trained model. To avoid the overfitting issues of the standard softmax function, we adopt the one-vs-each softmax approximation along with its Pólya-Gamma augmentation (OVE-PG). We evaluate our method on several benchmark datasets and demonstrate that using the Bayesian principle for prompt learning is indeed a sensible choice. Code is available at the \url{https://github.com/ParkLabML/Bayesian_Principles_Improve_Prompt_Learning_In_Vision_Language_Models.}} }
Endnote
%0 Conference Paper %T Bayesian Principles Improve Prompt Learning In Vision-Language Models %A Mingyu Kim %A Jongwoo Ko %A Mijung Park %B Proceedings of The 28th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2025 %E Yingzhen Li %E Stephan Mandt %E Shipra Agrawal %E Emtiyaz Khan %F pmlr-v258-kim25c %I PMLR %P 4078--4086 %U https://proceedings.mlr.press/v258/kim25c.html %V 258 %X Prompt learning is a popular fine-tuning method for vision-language models due to its efficiency. It requires a small number of additional learnable parameters while significantly enhancing performance on target tasks. However, most existing methods suffer from overfitting to fine-tuning data, yielding poor generalizability. To address this, we propose a new training objective function based on a Bayesian learning principle to balance adaptability and generalizability. We derive a prior over the logits, where the mean function is parameterized by the pre-trained model, while the posterior corresponds to the fine-tuned model. This objective establishes a balance by allowing the fine-tuned model to adapt to downstream tasks while remaining close to the pre-trained model. To avoid the overfitting issues of the standard softmax function, we adopt the one-vs-each softmax approximation along with its Pólya-Gamma augmentation (OVE-PG). We evaluate our method on several benchmark datasets and demonstrate that using the Bayesian principle for prompt learning is indeed a sensible choice. Code is available at the \url{https://github.com/ParkLabML/Bayesian_Principles_Improve_Prompt_Learning_In_Vision_Language_Models.}
APA
Kim, M., Ko, J. & Park, M.. (2025). Bayesian Principles Improve Prompt Learning In Vision-Language Models. Proceedings of The 28th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 258:4078-4086 Available from https://proceedings.mlr.press/v258/kim25c.html.

Related Material