# Continuous Prompt Generation from Linear Combination of Discrete Prompt Embeddings

arXiv: https://arxiv.org/abs/2312.10323 · HTML: https://mlsanity.com/paper/2312.10323/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 2 of its 5 references that are in the map.
- First posted on arXiv 16 Dec 2023.

## Builds on (most cited first, 2 of 2)

- [The Power of Scale for Parameter-Efficient Prompt Tuning](https://mlsanity.com/paper/2104.08691/) (2021)
- [Prompt Waywardness: The Curious Case of Discretized Interpretation of Continuous Prompts](https://mlsanity.com/paper/2112.08348/) (2021)

## Built on by (most cited first, 2 of 2)

- [Dynamic Hedging Strategies in Derivatives Markets with LLM-Driven Sentiment and News Analytics](https://mlsanity.com/paper/2504.04295/) (2025)
- [Understanding Visual Concepts Across Models](https://mlsanity.com/paper/2406.07506/) (2024)

## Abstract (arXiv)

The wayward quality of continuous prompts stresses the importance of their interpretability as unexpected and unpredictable behaviors appear following training, especially in the context of large language models automating people-sensitive tasks such as resume screening. In this paper we present a novel method of constructing continuous prompts via discrete prompt embeddings and evaluate improvements to continuous prompt interpretability and inference accuracy. For a set of manually designed discrete prompts D, which we tokenize and embed each into tensor form, we train a model to predict the weights such that the linear combinations of those prompts correspond to higher performance on natural language understanding tasks.