LanguaShrink:Reducing Token Overhead with Psycholinguistics
Abstract: As large language models (LLMs) improve their capabilities in handling complex tasks, the issues of computational cost and efficiency due to long prompts are becoming increasingly prominent. To accelerate model inference and reduce costs, we propose an innovative prompt compression framework called LanguaShrink. Inspired by the observation that LLM performance depends on the density and position of key information in the input prompts, LanguaShrink leverages psycholinguistic principles and the Ebbinghaus memory curve to achieve task-agnostic prompt compression. This effectively reduces prompt length while preserving essential information. We referred to the training method of OpenChat.The framework introduces part-of-speech priority compression and data distillation techniques, using smaller models to learn compression targets and employing a KL-regularized reinforcement learning strategy for training.\cite{wang2023openchat} Additionally, we adopt a chunk-based compression algorithm to achieve adjustable compression rates. We evaluate our method on multiple datasets, including LongBench, ZeroScrolls, Arxiv Articles, and a newly constructed novel test set. Experimental results show that LanguaShrink maintains semantic similarity while achieving up to 26 times compression. Compared to existing prompt compression methods, LanguaShrink improves end-to-end latency by 1.43 times.
Paper Type: Long
Research Area: Efficient/Low-Resource Methods for NLP
Research Area Keywords: Human-Centered NLP,Efficient/Low-Resource Methods for NLP
Contribution Types: NLP engineering experiment
Languages Studied: English,Chinese
Submission Number: 1084
Loading