Experiments in Agentic AI for Science

2026-05-25

AI Summary

This work addresses the limitations of large language models in scientific workflows—particularly their constrained context length and reasoning capabilities, which hinder autonomous execution of complex research tasks—by proposing a “Local Body/Remote Brain” hybrid architecture. This framework integrates local executors with cloud-based large models to develop two scientific agents: DeepTS/DeepCollector for automated large-scale time-series data processing, and DeepScribe for transforming complex mathematical and physical lectures into structured reports. The system leverages Cellular RAG for fine-grained knowledge extraction, remote data validation, and distributed concurrency control. Deployed via a Python coordinator in Google Colab, it has been successfully applied to time-series analysis, knowledge graph construction, and extended to high-energy physics (DeepQCD), significantly advancing the automation and intelligence of scientific research tasks.

Abstract

This paper details two novel frameworks for developing autonomous, agentic AI in scientific workflows. Both systems leverage a hybrid Local Body, Remote Brain architecture via Google Colab, utilizing Python-based local orchestrators to invoke large language model (LLM) cloud backends. The first agent, DeepTS/DeepCollector, automates the large-scale curation, extraction, and deduplication of time-series datasets. The second, DeepScribe, is an autonomous presentation analyzer that converts visually dense, mathematically complex physics lectures into structured scientific reports. Through practical systems engineering-such as granular attribute extraction (Cellular RAG), remote data inspection, and distributed concurrency controls-we demonstrate how agentic AI can overcome the context and reasoning limitations of current state-of-the-art systems to rigorously support scientific workflows. Finally, we outline a generalization of DeepTS to support deep knowledge graphs and discuss the application of this conceptual approach to high-energy physics (DeepQCD).

Problem

• Agentic AI
• Scientific Workflows
• Time-Series Data Curation
• Autonomous Presentation Analysis
• LLM Limitations

Innovation

• Agentic AI
• Local Body Remote Brain
• Cellular RAG
• Autonomous Scientific Workflow
• Deep Knowledge Graph