LLM Online Spatial-temporal Signal Reconstruction Under Noise

TL;DR

The LLM-OSR framework is introduced, which integrates Graph Signal Processing (GSP) and Large Language Models (LLMs) for online spatial-temporal signal reconstruction and demonstrates that utilizing GPT-4-o mini within the LLM-OSR is accurate and robust under Gaussian noise conditions.

Abstract

This work introduces the LLM Online Spatial-temporal Reconstruction (LLM-OSR) framework, which integrates Graph Signal Processing (GSP) and Large Language Models (LLMs) for online spatial-temporal signal reconstruction. The LLM-OSR utilizes a GSP-based spatial-temporal signal handler to enhance graph signals and employs LLMs to predict missing values based on spatiotemporal patterns. The performance of LLM-OSR is evaluated on traffic and meteorological datasets under varying Gaussian noise levels. Experimental results demonstrate that utilizing GPT-4-o mini within the LLM-OSR is accurate and robust under Gaussian noise conditions. The limitations are discussed along with future research insights, emphasizing the potential of combining GSP techniques with LLMs for solving spatiotemporal prediction tasks.

Figures (5)

• Figure 1: An overview of the LLM-OSR workflow
• Figure 2: The training process of the GSP-based spatial-temporal signal handler.
• Figure 3: The LLM-based Spatial-temporal Signal Predictor.
• Figure 4: The prompts prepared for LLM and the responses generated by the LLM (GPT-4o mini).
• Figure 5: The Seattle loop dataset at 4 different time instances.