AI Paper Atlas

Research map / Large language models / Language, safety and interpretability

InfoPrompt: Information-Theoretic Soft Prompt Tuning for Natural Language Understanding

Junda Wu, Tong Yu, Rui Wang, Zhao Song, Ruiyi Zhang, Handong Zhao, Chaochao Lu, Shuai Li, Ricardo Henao

NeurIPS 2023 · Large language models › Language, safety and interpretability · approach: Faithfulness of LLM explanations · 4 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 head loss (gamma=0.05) (ablation)InfoPrompt · GLUE full dataaverage score: 0.7531 -> 0.7590Improved
add representation loss (beta=0.1) (ablation)InfoPrompt · GLUE full dataaverage score: 0.7306 -> 0.7590Improved
add both info losses to naive prompt tuning (ablation)InfoPrompt · GLUE full dataaverage score: 0.7202 -> 0.7590Improved
add head loss (ablation)InfoPrompt · RE/NER/SemEval full dataaverage score: 0.7770 -> 0.7916Improved
add representation loss (ablation)InfoPrompt · RE/NER/SemEval full dataaverage score: 0.7746 -> 0.7916Improved
add both losses (ablation)InfoPrompt · RE/NER/SemEval full dataaverage score: 0.7519 -> 0.7916Improved
add head loss (ablation)InfoPrompt · GLUE few-shotaverage score: 0.5402 -> 0.5559Improved
add representation loss (ablation)InfoPrompt · GLUE few-shotaverage score: 0.5092 -> 0.5559Improved
add both losses (ablation)InfoPrompt · GLUE few-shotaverage score: 0.4925 -> 0.5559Improved
add head loss (ablation)InfoPrompt · RE/NER/SemEval few-shotaverage score: 0.3126 -> 0.3222Improved
add representation loss (ablation)InfoPrompt · RE/NER/SemEval few-shotaverage score: 0.3069 -> 0.3222Improved
use InfoPrompt instead of WARPsoft prompt tuning · RE/NER/SemEval few-shotaverage score: 0.2801 -> 0.3222Improved

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