Paper Detail

Bayesian Adaptation Gym: A Benchmark for the Bayesian Low-Rank Adaptation of Multi-Modal Language Models

Colin Samplawski, Ramneet Kaur, Manoj Acharya, Anirban Roy, Adam D. Cobb

Browse

Workflow Queues

arxiv Score 9.2

Published 2026-06-20 · First seen 2026-06-24

General AI

Open paper source

Abstract

Large multi-modal language models are increasingly deployed in high-stakes domains, making well-calibrated uncertainty essential. Traditional Bayesian methods approximate posteriors over all model weights, which becomes intractable for modern large models. For this reason, recent work instead considers Bayesian low-rank adaptation to enable tractable posterior approximation. Due to a lack of a standardized benchmark to evaluate these approaches, it remains unclear where these methods provide meaningful benefits. To fill this gap, we introduce Bayesian Adaptation Gym (BAG), a benchmark for the Bayesian adaptation of multi-modal language models. BAG provides reference implementations of classic Bayesian baselines and state-of-the-art adaptation methods, along with a multi-modal dataset and task suite designed to probe calibration, robustness under distribution shift, and decision-making under uncertainty via active learning. Using BAG, we conduct and report extensive experiments across model sizes, datasets, and tasks to highlight the successes and failures of current Bayesian adaptation approaches. To enable further research, BAG is fully open source: https://github.com/SRI-CSL/BayesAdapt.

Workflow Status

Review status
pending
Role
unreviewed
Read priority
soon
Vote
Not set.
Saved
no
Collections
Not filed yet.
Next action
Not filled yet.

Reading Brief

No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.

Why It Surfaced

No ranking explanation is available yet.

Tags

No tags.

Similar Papers

TalkLoRA: Communication-Aware Mixture of Low-Rank Adaptation for Large Language Models 2026-04-07 · Score 5.3 General AI TLoRA: Task-aware Low Rank Adaptation of Large Language Models 2026-04-20 · Score 9.3 General AI Bayesian Sparse Low-Rank Adaptation for Large Language Model Uncertainty Estimation 2026-07-02 · Score 12.6 General AI Cross-Modal Bayesian Low-Rank Adaptation for Uncertainty-Aware Multimodal Learning 2026-04-17 · Score 5.3 General AI Low-Rank Adaptation Redux for Large Models 2026-04-23 · Score 3.3 General AI

BibTeX

@article{samplawski2026bayesian,
  title = {Bayesian Adaptation Gym: A Benchmark for the Bayesian Low-Rank Adaptation of Multi-Modal Language Models},
  author = {Colin Samplawski and Ramneet Kaur and Manoj Acharya and Anirban Roy and Adam D. Cobb},
  year = {2026},
  abstract = {Large multi-modal language models are increasingly deployed in high-stakes domains, making well-calibrated uncertainty essential. Traditional Bayesian methods approximate posteriors over all model weights, which becomes intractable for modern large models. For this reason, recent work instead considers Bayesian low-rank adaptation to enable tractable posterior approximation. Due to a lack of a standardized benchmark to evaluate these approaches, it remains unclear where these methods provide mea},
  url = {https://arxiv.org/abs/2606.22188},
  keywords = {cs.LG},
  eprint = {2606.22188},
  archiveprefix = {arXiv},
}

Metadata

{}