AI Paper Atlas

Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes

Bayesian-LoRA: Probabilistic Low-Rank Adaptation of Large Language Models

Moule Lin, Shuhao Guan, Andrea Patane, David Gregg, Goetz Botterweck

ICML 2026 · Probabilistic and scientific modeling › Uncertainty and Gaussian processes · approach: Nonparametric and PAC-Bayes Bayesian learning · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add normalizing flow to the inducing posterior (L=1 vs L=0) (ablation)Bayesian-LoRA · OBQAaccuracy: 79.0 -> 81.6Improved
add normalizing flow to the inducing posterior (L=1 vs L=0) (ablation)Bayesian-LoRA · OBQANLL: 0.58 -> 0.49Improved
increase flow depth from 1 to 4 (ablation)Bayesian-LoRA · OBQAECE: 5.7 -> 4.9Improved
increase flow depth from 1 to 2 (ablation)Bayesian-LoRA · OBQAaccuracy: 81.6 -> 80.8Hurt
increase flow depth from 1 to 4 (ablation)Bayesian-LoRA · OBQAtrain time (x MAP): 1.23 -> 1.38Hurt
increase inducing dimension r=c (ablation)Bayesian-LoRA · OBQAECEImproved
add full probabilistic posterior over degenerate (MAP-limit) setting (ablation)Bayesian-LoRA · OBQAaccuracy: 77.81 -> 81.60Improved
add full probabilistic posterior over degenerate (MAP-limit) setting (ablation)Bayesian-LoRA · OBQAECE: 9.77 -> 5.70Improved
use Bayesian-LoRA instead of BLoBLlama-2-7B · WinoGrande-Saccuracy: 69.07 -> 70.90Improved
use Bayesian-LoRA instead of LoRA MAPLlama-2-7B · WinoGrande-SECE: 30.80 -> 4.90Improved
use Bayesian-LoRA N=2 instead of LoRA MAPLlama-2-7B · WikiText-2 top 5% entropy tokensECE (test): 1.60 -> 1.26Improved
use Bayesian-LoRA instead of LoRA baseline FTQwen2.5-14B-Instruct · MATHCoT-NLL: 2.165 -> 0.513Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy