LLM-Collab: a framework for enhancing task planning via chain-of-thought and multi-agent collaboration DOI

Hong Phong Cao,

Rong Ma, Yanlong Zhai

et al.

Applied Computing and Intelligence, Journal Year: 2024, Volume and Issue: 4(2), P. 328 - 348

Published: Jan. 1, 2024

<p>Large language models have shown strong capabilities in performing natural planning tasks, largely due to the chain-of-thought method, which enhances their ability solve complex tasks through explicit intermediate inference. However, they face challenges acquiring new knowledge, executing calculations, and interacting with environment. Although previous work has enabled large use external tools improve reasoning environmental interaction, there was no scalable or cohesive structure for these technologies. In this paper, we present LLM-Collab, where Collab represents cooperative interaction between two AI agents, model plays a key role creation of agents. For took as core agents designed cooperate on tasks: One an analyst tool selection phase validation, other executor specific tasks. Our method provided comprehensive list facilitate invocation integration ensuring seamless collaboration process. This paradigm established unified framework autonomous task-solving based massive by demonstrating how communication enable multi-agent collaboration.</p>

Language: Английский

High Heels, Compass, Spider-Man, or Drug? Metaphor Analysis of Generative Artificial Intelligence in Academic Writing DOI Creative Commons
Fangzhou Jin, Lanfang Sun, Y. C. Pan

et al.

Computers & Education, Journal Year: 2025, Volume and Issue: unknown, P. 105248 - 105248

Published: Jan. 1, 2025

Language: Английский

Citations

6
Modifying AI, Enhancing Essays: How Active Engagement with Generative AI Boosts Writing Quality DOI
Kaixun Yang, Mladen Raković, Zhiping Liang

et al.

Published: Feb. 21, 2025

Language: Английский

Citations

1
A Genre, Scoring, and Authorship Analysis of AI-Generated and Human-Written Refusal Emails DOI

Willie Wilson,

Heath Rose

Business and Professional Communication Quarterly, Journal Year: 2025, Volume and Issue: unknown

Published: March 12, 2025

This study compares AI-generated (ChatGPT and Gemini) human-written business refusal texts. A genre analysis found that texts are formulaic less nuanced than Applying a rating of professional writing quality, inferential statistics revealed no significant difference in scores between Gemini texts, but ChatGPT as lower. Human assessors identified authorship with an accuracy rate 68.1%, 86% accuracy. Key concerns for were tone, relationship, language choice, content, structure. The findings inform four key areas focus teaching the AI age.

Language: Английский

Citations

0
Not Just Novelty: A Longitudinal Study on Utility and Customization of an AI Workflow DOI
Tao Long, Katy Ilonka Gero, Lydia B. Chilton

et al.

Designing Interactive Systems Conference, Journal Year: 2024, Volume and Issue: unknown, P. 782 - 803

Published: June 29, 2024

Language: Английский

Citations

0
LLM-Collab: a framework for enhancing task planning via chain-of-thought and multi-agent collaboration DOI

Hong Phong Cao,

Rong Ma, Yanlong Zhai

et al.

Applied Computing and Intelligence, Journal Year: 2024, Volume and Issue: 4(2), P. 328 - 348

Published: Jan. 1, 2024

<p>Large language models have shown strong capabilities in performing natural planning tasks, largely due to the chain-of-thought method, which enhances their ability solve complex tasks through explicit intermediate inference. However, they face challenges acquiring new knowledge, executing calculations, and interacting with environment. Although previous work has enabled large use external tools improve reasoning environmental interaction, there was no scalable or cohesive structure for these technologies. In this paper, we present LLM-Collab, where Collab represents cooperative interaction between two AI agents, model plays a key role creation of agents. For took as core agents designed cooperate on tasks: One an analyst tool selection phase validation, other executor specific tasks. Our method provided comprehensive list facilitate invocation integration ensuring seamless collaboration process. This paradigm established unified framework autonomous task-solving based massive by demonstrating how communication enable multi-agent collaboration.</p>

Language: Английский

Citations

0