AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Multitask Prompt Tuning Enables Parameter-Efficient Transfer Learning

Zhen Wang, Rameswar Panda, Leonid Karlinsky, Rogerio Feris, Huan Sun, Yoon Kim

ICLR 2023 · Label-efficient and robust learning › Recognition and long-tail · approach: Multilingual language representation pretraining · 30 citations (OpenAlex)

Read the paper

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
transfer a shared prompt learned by multitask prompt decomposition and distillation instead of randomly initialized promptsprompt tuning · GLUEaverage score: 13%Improved
transfer a shared prompt learned by multitask prompt decomposition and distillation instead of randomly initialized promptsprompt tuning · SuperGLUEaverage score: 16%Improved
use a shared transferable prompt with low-rank task-specific updates rather than vanilla prompt tuningprompt tuning · MRQAaverage F1: +2.8%Improved
use a shared transferable prompt with low-rank task-specific updates rather than vanilla prompt tuningprompt tuning · Othersaverage accuracy: +13.5%Improved
add prompt decomposition into shared and task-specific components (ablation)MPT source prompts without distillation · SuperGLUEaverage score: 69.5 -> 73.0Improved
add prompt distillation from separately trained source prompts (ablation)shared vanilla source prompt · SuperGLUEaverage score: 69.5 -> 70.6Improved
combine prompt decomposition with prompt distillation (ablation)MPT source prompts · SuperGLUEaverage score: 74.1Improved
add hidden-state matching to logits-based distillation (ablation)MPT distillation objective · SuperGLUEaverage score: 73.7% vs. 74.1%Improved
use logits and hidden-state distillation instead of direct prompt-distance MSE (ablation)MPT distillation objective · SuperGLUEaverage score: 73.6% vs. 74.1%Improved
increase prompt length from 100 to 300 (ablation)MPT · SuperGLUEaccuracy: 76.8%Improved
increase prompt length from 300 to 400 (ablation)MPT · SuperGLUEaccuracy: absolute 1.8% dropHurt
update both the shared component and task-specific vectors during adaptation (ablation)MPT target adaptation · SuperGLUEaverage score: 62.5% and 71.3%Improved

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