A learned translator converts soft prompts to hard natural-language prompts, outperforming training-free baselines like InSPEcT and enabling portable prompts that exceed original soft-prompt performance on larger models.
InProceedings of the 52nd An- nual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 1566– 1576, Baltimore, Maryland
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Learning to Translate from Soft to Hard LLM Prompts
A learned translator converts soft prompts to hard natural-language prompts, outperforming training-free baselines like InSPEcT and enabling portable prompts that exceed original soft-prompt performance on larger models.