GeoTool-GPT: a trainable method for facilitating Large Language Models to master GIS tools

International Journal of Geographical Information Science, Journal Year: 2024, Volume and Issue: 39(4), P. 707 - 731

Published: Dec. 11, 2024

Large Language Models (LLMs) excel in natural language-relevant tasks like text generation and question answering Q&A. To further expand their application, efforts focus on enabling LLMs to utilize real-world tools. However, tool-use ability professional GIS remains under explored due two main challenges. Firstly, are usually trained general-domain corpora, lacking sufficient comprehensive GIS-specific data align with knowledge, including understanding the functions of Secondly, researchers often need combine multiple tools solve geospatial tasks. address these challenges, we propose a trainable method enable master We curated set resources: instruction-response (GeoTool, 1950 instructions) enhance for tools, instruction-solution (GeoSolution, 3645 improve generate solutions tasks, annotated evaluation (GeoTask, 300 evaluating LLMs' proficiency. Using collected training (GeoTool GeoSolution), fine-tuned professional-domain LLM called GeoTool-GPT based an open-source LLM, LLaMA-2-7b model. The experiment validates our method's effectiveness enhancing domain, performance model closely approaching that GPT-4.

Language: Английский

---

Large Language Model-Empowered Agents for Simulating Macroeconomic Activities

SSRN Electronic Journal, Journal Year: 2023, Volume and Issue: unknown

Published: Jan. 1, 2023

The advent of the Web has brought about a paradigm shift in traditional economics, particularly digital economy era, enabling precise recording and analysis individual economic behavior. This led to growing emphasis on data-driven modeling macroeconomics. In macroeconomic research, Agent-based (ABM) emerged as an alternative, evolving through rule-based agents, machine learning-enhanced decision-making, and, more recently, advanced AI agents. However, existing works are suffering from three main challenges when endowing agents with human-like including agent heterogeneity, influence trends, multifaceted factors. Large language models (LLMs) have recently gained prominence offering autonomous characteristics. Therefore, leveraging LLMs simulation presents opportunity overcome limitations. this work, we take early step introducing novel approach that leverages simulation. We design prompt-engineering-driven LLM exhibit decision-making adaptability environment, abilities perception, reflection, address abovementioned challenges. Simulation experiments activities show LLM-empowered can make realistic work consumption decisions emerge reasonable phenomena than or Our demonstrates promising potential simulate macroeconomics based its

Language: Английский

---

Homogenizing Effect of Large Language Model (LLM) on Creative Diversity: An Empirical Comparison of Human and ChatGPT Writing

Published: Aug. 22, 2024

Generative AI systems, especially Large Language Models (LLMs) like ChatGPT, have recently emerged as significant contributors to creative processes. While LLMs can produce content that might be good or even better than human creations, their widespread use risks reducing the diversity of outputs across groups people. In present research, we aimed quantify this homogenizing effect LLMS on collective creativity. Across three preregistered studies, analyzed 2,200 college admissions essays. Using a novel measure—diversity growth rate—we showed each additional human-written essay contributed more new ideas GPT-4 essay. This persisted after range enhancements writings, including prompt and parameters modifications. Overall, our findings suggest that, despite improvements in individual creativity, could diminish ideas.

Language: Английский

---

Player-Driven Emergence in LLM-Driven Game Narrative

2021 IEEE Conference on Games (CoG), Journal Year: 2024, Volume and Issue: unknown, P. 1 - 8

Published: Aug. 5, 2024

Language: Английский

---

Exploring the Synergy of Grammar-Aware Prompt Engineering and Formal Methods for Mitigating Hallucinations in LLMs

East African Journal of Information Technology, Journal Year: 2024, Volume and Issue: 7(1), P. 188 - 201

Published: Aug. 15, 2024

Recent advancements in Artificial Intelligence (AI), particularly the advanced machine learning for Natural Language Processing (NLP) paradigm, have led to development of powerful Large Models (LLMs) capable impressive feats tasks like translation, text summarisation, generation and code generation. However, a critical challenge hindering their real-world deployment is susceptibility hallucinations, where they generate plausible looking but factually incorrect outputs. These limitations come with adverse effects, such as propagation misinformation reducing user trustworthiness related technologies, even when possess transformative potential various sectors. This study aims enhance performance LLMs by presenting new strategy that combines grammar-aware prompt engineering (GAPE) formal methods (FMs) leverage synergy LLM process logic. We argue combining linguistic principles using GAPE constructing basis structures FMs, we could improve LLM's ability analyse language, decrease ambiguity prompts, consistency output, eventually, greatly diminish hallucinations. To do this, propose collaboration between linguists AI experts while also providing specialised training emphasises precision. Additionally, suggest implementing iterative design procedures use FM continuously LLMs. By following these techniques, may create future which are more trustworthy wide range users cases reliable technologies efficient practical situations

Language: Английский

---

SARD: A Human-AI Collaborative Story Generation

Lecture notes in computer science, Journal Year: 2024, Volume and Issue: unknown, P. 94 - 105

Published: Dec. 16, 2024

Language: Английский

---

Beyond the Fourth Paradigm — the Rise of AI

Published: Sept. 25, 2023

Thanks to the availability of huge amounts data and improved computational resources, AI methods are gaining importance in scientific workflows, from image recognition natural language processing materials science. In many domains usage is under active investigation first results show a tremendous potential, suggesting that will have significant impact way beyond currently dominating examples processing.

Language: Английский

---