Large Language Models Can Infer Personality from Free-Form User Interactions DOI Open Access
Heinrich Peters, Moran Cerf, Sandra Matz

et al.

Published: May 20, 2024

This study investigates the capacity of Large Language Models (LLMs) to infer Big Five personality traits from free-form user interactions. The results demonstrate that a chatbot powered by GPT-4 can with moderate accuracy, outperforming previous approaches drawing inferences static text content. accuracy varied across different conversational settings. Performance was highest when prompted elicit personality-relevant information users (mean r=.443, range=[.245, .640]), followed condition placing greater emphasis on naturalistic interaction r=.218, range=[.066, .373]). Notably, direct focus assessment did not result in less positive experience, participants reporting interactions be equally natural, pleasant, engaging, and humanlike both conditions. A mimicking ChatGPT’s default behavior acting as helpful assistant led markedly inferior lower experience ratings but still captured psychologically meaningful for some r=.117, range=[-.004, .209]). Preliminary analyses suggest varies only marginally socio-demographic subgroups. Our highlight potential LLMs psychological profiling based We discuss practical implications ethical challenges associated these findings.

Language: Английский

A survey on large language model based autonomous agents DOI Creative Commons
Lei Wang, Chen Ma, Xueyang Feng

et al.

Frontiers of Computer Science, Journal Year: 2024, Volume and Issue: 18(6)

Published: March 22, 2024

Abstract Autonomous agents have long been a research focus in academic and industry communities. Previous often focuses on training with limited knowledge within isolated environments, which diverges significantly from human learning processes, makes the hard to achieve human-like decisions. Recently, through acquisition of vast amounts Web knowledge, large language models (LLMs) shown potential human-level intelligence, leading surge LLM-based autonomous agents. In this paper, we present comprehensive survey these studies, delivering systematic review holistic perspective. We first discuss construction agents, proposing unified framework that encompasses much previous work. Then, overview diverse applications social science, natural engineering. Finally, delve into evaluation strategies commonly used for Based also several challenges future directions field.

Language: Английский

Citations

215
The rise and potential of large language model based agents: a survey DOI

Zhiheng Xi,

Wen-Xiang Chen, Xin Hua Guo

et al.

Science China Information Sciences, Journal Year: 2025, Volume and Issue: 68(2)

Published: Jan. 17, 2025

Language: Английский

Citations

9
Assessing the Alignment of Large Language Models With Human Values for Mental Health Integration: Cross-Sectional Study Using Schwartz’s Theory of Basic Values DOI Creative Commons
Dorit Hadar‐Shoval, Kfir Asraf, Yonathan Mizrachi

et al.

JMIR Mental Health, Journal Year: 2024, Volume and Issue: 11, P. e55988 - e55988

Published: March 8, 2024

Large language models (LLMs) hold potential for mental health applications. However, their opaque alignment processes may embed biases that shape problematic perspectives. Evaluating the values embedded within LLMs guide decision-making have ethical importance. Schwartz's theory of basic (STBV) provides a framework quantifying cultural value orientations and has shown utility examining in contexts, including cultural, diagnostic, therapist-client dynamics.

Language: Английский

Citations

15
The use of ChatGPT for personality research: Administering questionnaires using generated personas DOI Creative Commons
Joost de Winter, Tom Driessen, Dimitra Dodou

et al.

Personality and Individual Differences, Journal Year: 2024, Volume and Issue: 228, P. 112729 - 112729

Published: June 3, 2024

Personality research has traditionally relied on questionnaires, which bring with them inherent limitations, such as response style bias. With the emergence of large language models ChatGPT, question arises to what extent these can be used in personality research. In this study, ChatGPT (GPT-4) generated 2000 text-based personas. Next, for each persona, completed a short form Big Five Inventory (BFI-10), Brief Sensation Seeking Scale (BSSS), and Short Dark Triad (SD3). The mean scores BFI-10 items were found correlate strongly means from previously published research, principal component analysis revealed clear five-component structure. Certain relationships between traits, negative correlation age persona BSSS score, clearly interpretable, while some other correlations diverged literature. An additional using four new sets personas each, including set 'realistic' cinematic personas, showed that matrix among constructs was affected by set. It is concluded evaluating questionnaires hypotheses prior engaging real individuals holds promise.

Language: Английский

Citations

10
A survey on LLM-based multi-agent systems: workflow, infrastructure, and challenges DOI Creative Commons
Xinyi Li, S. Wang, Siqi Zeng

et al.

Vicinagearth., Journal Year: 2024, Volume and Issue: 1(1)

Published: Oct. 8, 2024

Abstract The pursuit of more intelligent and credible autonomous systems, akin to human society, has been a long-standing endeavor for humans. Leveraging the exceptional reasoning planning capabilities large language models (LLMs), LLM-based agents have proposed achieved remarkable success across wide array tasks. Notably, multi-agent systems (MAS) are considered promising pathway towards realizing general artificial intelligence that is equivalent or surpasses human-level intelligence. In this paper, we present comprehensive survey these studies, offering systematic review MAS. Adhering workflow synthesize structure encompassing five key components: profile, perception, self-action, mutual interaction, evolution. This unified framework encapsulates much previous work in field. Furthermore, illuminate extensive applications MAS two principal areas: problem-solving world simulation. Finally, discuss detail several contemporary challenges provide insights into potential future directions domain.

Language: Английский

Citations

8
Generative AI from Theory to Practice: A Case Study of Financial Advice DOI Creative Commons

Andrew W. Lo,

J. Perran Ross

Published: March 27, 2024

We identify some of the most pressing issues facing adoption large language models (LLMs) in practical settings and propose a research agenda to reach next technological inflection point generative AI. focus on three challenges LLM applications: domain-specific expertise ability tailor that user's unique situation, trustworthiness adherence moral ethical standards, conformity regulatory guidelines oversight.

Language: Английский

Citations

7
Bias and Unfairness in Information Retrieval Systems: New Challenges in the LLM Era DOI Creative Commons
Sunhao Dai, Xu Chen, Shicheng Xu

et al.

Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, Journal Year: 2024, Volume and Issue: unknown, P. 6437 - 6447

Published: Aug. 24, 2024

With the rapid advancements of large language models (LLMs), information retrieval (IR) systems, such as search engines and recommender have undergone a significant paradigm shift. This evolution, while heralding new opportunities, introduces emerging challenges, particularly in terms biases unfairness, which may threaten ecosystem. In this paper, we present comprehensive survey existing works on pressing bias unfairness issues IR systems when integration LLMs. We first unify distribution mismatch problems, providing groundwork for categorizing various mitigation strategies through alignment. Subsequently, systematically delve into specific arising from three critical stages LLMs systems: data collection, model development, result evaluation. doing so, meticulously review analyze recent literature, focusing definitions, characteristics, corresponding associated with these issues. Finally, identify highlight some open problems challenges future work, aiming to inspire researchers stakeholders field beyond better understand mitigate LLM era. also consistently maintain GitHub repository relevant papers resources rising direction at https://github.com/KID-22/LLM-IR-Bias-Fairness-Survey.

Language: Английский

Citations

6
Large language models can outperform humans in social situational judgments DOI Creative Commons

Justin Mittelstädt,

Julia Maier, Panja Goerke

et al.

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: Nov. 10, 2024

Abstract Large language models (LLM) have been a catalyst for the public interest in artificial intelligence (AI). These technologies perform some knowledge-based tasks better and faster than human beings. However, whether AIs can correctly assess social situations devise socially appropriate behavior, is still unclear. We conducted an established Situational Judgment Test (SJT) with five different chatbots compared their results responses of participants ( N = 276). Claude, Copilot you.com’s smart assistant performed significantly humans proposing suitable behaviors situations. Moreover, effectiveness rating behavior options aligned well expert ratings. indicate that LLMs are capable producing adept judgments. While this constitutes important requirement use as virtual assistants, challenges risks associated wide-spread contexts.

Language: Английский

Citations

5
Exploring people's perceptions of LLM-generated advice DOI Creative Commons
Joel Wester, S. de Jong, Henning Pohl

et al.

Computers in Human Behavior Artificial Humans, Journal Year: 2024, Volume and Issue: 2(2), P. 100072 - 100072

Published: June 7, 2024

When searching and browsing the web, more of information we encounter is generated or mediated through large language models (LLMs). This can be looking for a recipe, getting help on an essay, relationship advice. Yet, there limited understanding how individuals perceive advice provided by these LLMs. In this paper, explore people's perception LLM-generated advice, what role diverse user characteristics (i.e., personality technology readiness) play in shaping their perception. Further, as difficult to distinguish from human assess perceived creepiness such To investigate this, run exploratory study (N = 91), where participants rate different styles (generated GPT-3.5 Turbo). Notably, our findings suggest that who identify agreeable tend like find it useful. with higher technological insecurity are likely follow useful, deem friend could have given Lastly, see 'skeptical' style was rated most unpredictable, 'whimsical' least malicious—indicating LLM influence perceptions. Our results also provide overview considerations likelihood, receptiveness, they seek digital assistants. Based results, design takeaways outline future research directions further inform support applications targeting people expectations needs.

Language: Английский

Citations

4
Large language models can replicate cross-cultural differences in personality DOI
Paweł Niszczota,

Mateusz Janczak,

Michał Misiak

et al.

Journal of Research in Personality, Journal Year: 2025, Volume and Issue: unknown, P. 104584 - 104584

Published: Jan. 1, 2025

Citations

0