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
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add head loss (gamma=0.05) (ablation) | InfoPrompt · GLUE full data | average score: 0.7531 -> 0.7590 | Improved |
| add representation loss (beta=0.1) (ablation) | InfoPrompt · GLUE full data | average score: 0.7306 -> 0.7590 | Improved |
| add both info losses to naive prompt tuning (ablation) | InfoPrompt · GLUE full data | average score: 0.7202 -> 0.7590 | Improved |
| add head loss (ablation) | InfoPrompt · RE/NER/SemEval full data | average score: 0.7770 -> 0.7916 | Improved |
| add representation loss (ablation) | InfoPrompt · RE/NER/SemEval full data | average score: 0.7746 -> 0.7916 | Improved |
| add both losses (ablation) | InfoPrompt · RE/NER/SemEval full data | average score: 0.7519 -> 0.7916 | Improved |
| add head loss (ablation) | InfoPrompt · GLUE few-shot | average score: 0.5402 -> 0.5559 | Improved |
| add representation loss (ablation) | InfoPrompt · GLUE few-shot | average score: 0.5092 -> 0.5559 | Improved |
| add both losses (ablation) | InfoPrompt · GLUE few-shot | average score: 0.4925 -> 0.5559 | Improved |
| add head loss (ablation) | InfoPrompt · RE/NER/SemEval few-shot | average score: 0.3126 -> 0.3222 | Improved |
| add representation loss (ablation) | InfoPrompt · RE/NER/SemEval few-shot | average score: 0.3069 -> 0.3222 | Improved |
| use InfoPrompt instead of WARP | soft prompt tuning · RE/NER/SemEval few-shot | average score: 0.2801 -> 0.3222 | Improved |
Closest papers by meaning
- EPT: Efficient Prompt Tuning by Multi-Space Projection and Prompt Fusion (AAAI 2025 · similarity 0.90)
- BayesPrompt: Prompting Large-Scale Pre-Trained Language Models on Few-shot Inference via Debiased Domain Abstraction (ICLR 2024 · similarity 0.90)
- DePT: Decomposed Prompt Tuning for Parameter-Efficient Fine-tuning (ICLR 2024 · similarity 0.89)
- All You Need is One: Capsule Prompt Tuning with a Single Vector (NeurIPS 2025 · similarity 0.89)
- ADePT: Adaptive Decomposed Prompt Tuning for Parameter-Efficient Fine-tuning (ICLR 2025 · similarity 0.89)
- Evaluating the Zero-shot Robustness of Instruction-tuned Language Models (ICLR 2024 · similarity 0.88)
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.
