Latest Articles for Agents

Examining how strategic interactions impact machine learning model performance.

A method for agents to estimate functions without sharing data directly.

GPT-4 demonstrates its ability to engage with the classic game DOOM.

Addressing value overestimation and primacy bias to enhance agent performance.

A new approach to understanding crypto market dynamics through advanced modeling techniques.

Explore solutions for agents moving through shared spaces without collisions.

Exploring how deep learning helps us study the emergence of language.

Optimizing agent behavior through norms and sanctions in complex systems.

MA4DIV enhances online search results by improving document diversity through a multi-agent approach.

A new method improves self-driving car training through human-like interaction.

New method improves agent collaboration in diverse learning environments while maintaining data privacy.

A new framework aiding research in the complex art market.

Study explores how language models work with humans and each other in task completion.

A system that helps robots detect and fix task failures in real-time.

A look into how agents adapt their learning for food gathering.

Examining how agents and principals share information on performance metrics.

A framework using agents empowers users to manage IoT security and privacy preferences.

Mode consensus algorithms streamline agreement in multi-agent systems efficiently.

Explore how basic rules can guide agents through mazes with minimal information.

A focus on methods to assess and refine digital agents' performance.

This study addresses challenges in bidding strategies using artificial currencies.

This work presents a strategy for agents to independently diagnose and resolve issues.

Exploring effective mechanisms for placing service facilities based on user demand.

A system that simulates battles to reveal soldiers' experiences.

This article discusses strategies to find black holes in networks.

EasyLAN streamlines the development of collaborative LLM networks for complex tasks.

Discover how Recursive Backwards Q-Learning improves decision-making in AI.

A method for agents to maintain shapes with minimal communication.

This article examines trust issues in robot teams for reliable data gathering.

Learn how groups of agents can agree despite unreliable members.

Explore economic principles through simple scientific ideas for better insight.

Exploring how trust develops in human-agent interactions over time.

A new method improves machine learning training efficiency while protecting data privacy.

WorkBench tests agents' ability to perform realistic office tasks with a unique evaluation method.

Discover how perception engineering creates believable experiences through sensory manipulation.

A new method ensures privacy while allowing efficient resource sharing among agents.

New algorithm enhances learning in real-world tasks without resets.

Introducing T-PoP, a method for privacy-focused location proof.

Agents make choices in a multi-agent setting using predictions for better outcomes.

A fresh method enhances agent adaptability across varied environments without specific tasks.