Leveraging Large Language Models and Agent-Based Systems for Scientific Data Analysis: Validation Study
Bibliographic record
Abstract
Background: Large language models have shown promise in transforming how complex scientific data are analyzed and communicated, yet their application to scientific domains remains challenged by issues of factual accuracy and domain-specific precision. The Laureate Institute for Brain Research-Tulsa University (LIBR-TU) Research Agent (LITURAt) leverages a sophisticated agent-based architecture to mitigate these limitations, using external data retrieval and analysis tools to ensure reliable, context-aware outputs that make scientific information accessible to both experts and nonexperts. Objective: The objective of this study was to develop and evaluate LITURAt to enable efficient analysis and contextualization of complex scientific datasets for diverse user expertise levels. Methods: An agent-based system based on large language models was designed to analyze and contextualize complex scientific datasets using a "plan-and-solve" framework. The system dynamically retrieves local data and relevant PubMed literature, performs statistical analyses, and generates comprehensive, context-aware summaries to answer user queries with high accuracy and consistency. Results: Our experiments demonstrated that LITURAt achieved an internal consistency rate of 94.8% and an external consistency rate of 91.9% across repeated and rephrased queries. Additionally, GPT-4 evaluations rated 80.3% (171/213) of the system's answers as accurate and comprehensive, with 23.5% (50/213) receiving the highest rating of 5 for completeness and precision. Conclusions: These findings highlight the potential of LITURAt to significantly enhance the accessibility and accuracy of scientific data analysis, achieving high consistency and strong performance in complex query resolution. Despite existing limitations, such as model stability for highly variable queries, LITURAt demonstrates promise as a robust tool for democratizing data-driven insights across diverse scientific domains.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.025 | 0.084 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.003 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".