Skip to content
Snippets Groups Projects
Commit 15f998b1 authored by Maxime MORGE's avatar Maxime MORGE
Browse files

LLM4AAMAS: Add fan24aaai

parent 6329f5c2
No related branches found
No related tags found
No related merge requests found
...@@ -416,6 +416,18 @@ A simulation of the propagation processes in a social network. ...@@ -416,6 +416,18 @@ A simulation of the propagation processes in a social network.
Lu, Jinzhu Mao, Jinghua Piao, Huandong Wang, Depeng Jin, Yong Li (2023)* Lu, Jinzhu Mao, Jinghua Piao, Huandong Wang, Depeng Jin, Yong Li (2023)*
Published on *arXiv* arXiv:2307.14984 Published on *arXiv* arXiv:2307.14984
This paper is a systematic analysis of the capabilities boundaries of LLMs in
social science. The authors select 3 classical games : dictator game,
Rock-Paper-Scissors and ring network game. They conclude that LLMs struggle to
build desires based on uncommon preferences, fail to refine belief from many
simple patterns, and may modify refined belief when taking actions.
- **[Can Large Language Models Serve as Rational Players in Game Theory? A
Systematic Analysis](https://ojs.aaai.org/index.php/AAAI/article/view/29751)**
Caoyun Fan, Jindou Chen, Yaohui Jin, and Hao He (2024) *Presented at AAAI*,
**38**(16), 17960-17967. [DOI:
10.1609/aaai.v38i16.29751](https://doi.org/10.1609/aaai.v38i16.29751)
When LLM-based agents participate in various games designed to assess different When LLM-based agents participate in various games designed to assess different
traits—such as the dictator game (altruism), the ultimatum game (fairness), the traits—such as the dictator game (altruism), the ultimatum game (fairness), the
trust game (trust, fairness, altruism, and reciprocity), the bomb risk game trust game (trust, fairness, altruism, and reciprocity), the bomb risk game
...@@ -426,12 +438,12 @@ occur, chatbots tend to be more cooperative and altruistic, displaying higher ...@@ -426,12 +438,12 @@ occur, chatbots tend to be more cooperative and altruistic, displaying higher
levels of trust, generosity, and reciprocity. They behave as if they prioritize levels of trust, generosity, and reciprocity. They behave as if they prioritize
maximizing the total payoff of both players rather than solely their own gain. maximizing the total payoff of both players rather than solely their own gain.
- **[A Turing test of whether AI chatbots are behaviorally similar to - **[A Turing test of whether AI chatbots are behaviorally similar to
humans](https://doi.org/10.1073/pnas.2313925121)** Qiaozhu Mei, Yutong Xie, humans](https://doi.org/10.1073/pnas.2313925121)** Qiaozhu Mei, Yutong Xie,
Walter Yuan, Matthew O. Jackson (2024) in Proceedings of the National Academy Walter Yuan, Matthew O. Jackson (2024) in Proceedings of the National Academy
of Sciences*, 121(9). of Sciences*, 121(9).
A study of LLMS as artificial social agents playing the iterated prisoner's A study of LLMS as artificial social agents playing the iterated prisoner's
dilemma which shows that prompt comprehension, memory representation, and dilemma which shows that prompt comprehension, memory representation, and
duration of the simulation play crucial roles. LLMs are characterized by an duration of the simulation play crucial roles. LLMs are characterized by an
...@@ -450,6 +462,8 @@ dilemma where aggressive strategies can persist or even dominate. ...@@ -450,6 +462,8 @@ dilemma where aggressive strategies can persist or even dominate.
Dilemma](https://arxiv.org/abs/2501.16173)** Richard Willis, Yali Du, Joel Z. Dilemma](https://arxiv.org/abs/2501.16173)** Richard Willis, Yali Du, Joel Z.
Leibo, Michael Luck (2025) Published on arXiv Leibo, Michael Luck (2025) Published on arXiv
### Generative MAS on the shelf ### Generative MAS on the shelf
- [MetaGPT](https://github.com/geekan/MetaGPT) is a framework for creating - [MetaGPT](https://github.com/geekan/MetaGPT) is a framework for creating
......
0% Loading or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment