Takaya Arita

Evolutionary stability of ordered bargaining signals under payoff inequality

Masanori Takano, Takaya Arita

jp (code pays fourni par la source)

Emergence of Social Reasoning through Adaptive Evolution of LLMs

Takumi Hara, Takaya Arita, Reiji Suzuki

Large language models (LLMs) exhibit advanced social reasoning capabilities like Theory of Mind (ToM), yet the dynamic acquisition process remains underexplored. We propose a constructive, artificial-life approach that treats LLMs as “model organisms,” investigating the emergent mechanisms of cognitive functions through biological …

jp (code pays fourni par la source)

An evolutionary model of animats with VLM-based subjective evaluation

Shota Miyazaki, Takaya Arita, Reiji Suzuki

In this study, we propose a framework that incorporates subjective evaluations provided by a Vision-Language Model (VLM) into the fitness evaluation and selection processes of a genetic algorithm. As the target of evolution, we employ virtual soft robots with flexible morphologies and …

An evolutionary model of animats with VLM-based subjective evaluation

Shota Miyazaki, Takaya Arita, Reiji Suzuki

jp (code pays fourni par la source)

Codes and data for "Emergence of Social Reasoning through Adaptive Evolution of LLMs"

Takumi Hara, Takaya Arita, Reiji Suzuki

Code and data for the ALIFE 2026 paper "Emergence of Social Reasoning through Adaptive Evolution of LLMs." Acknowledgment of Upstream Codebase:Our evolutionary computation framework is built upon GENOME (Zhang et al., 2025). We deeply appreciate the original authors for open-sourcing their excellent …

jp (code pays fourni par la source)

Data and code for "An evolutionary model of animats with VLM-based subjective evaluation"

Shota Miyazaki, Takaya Arita, Reiji Suzuki

This repository contains the data and source code associated with the study "An Evolutionary Model of Animats with VLM-Based Subjective Evaluation." The materials include experimental results from VLM-based subjective evaluation experiments, in which animats were evolved through pairwise comparisons performed by a …

jp (code pays fourni par la source)

Data and code for "An evolutionary model of animats with VLM-based subjective evaluation"

Shota Miyazaki, Takaya Arita, Reiji Suzuki

This repository contains the data and source code associated with the study "An Evolutionary Model of Animats with VLM-Based Subjective Evaluation." The materials include experimental results from VLM-based subjective evaluation experiments, in which animats were evolved through pairwise comparisons performed by a …

jp (code pays fourni par la source)

Emergence of Social Reasoning through Adaptive Evolution of LLMs

Takumi Hara, Takaya Arita, Reiji Suzuki

Code and data for the ALIFE 2026 paper "Emergence of SocialReasoning through Adaptive Evolution of LLMs." Initial minimal release; an organized version will be added later on the same record.

jp (code pays fourni par la source)

How memory can affect collective and cooperative behaviors in an LLM-Based Social Particle Swarm

Taisei Hishiki, Takaya Arita, Reiji SUZUKI

This study examines how memory shapes the collective and cooperative dynamics of Large Language Model (LLM) agents in a multi-agent system. To this end, we extend the Social Particle Swarm (SPS) model, in which agents move in a two-dimensional space and play …

How Reward and the Elimination of Dependence Thereon Lead Evolution to High-Bandwidth Learning

Solvi Arnold, Reiji Suzuki, Takaya Arita

Efficient learning is fundamental to intelligence. Popular AI learning methods still fall short of advanced biological learners in terms of efficiency, and hence, in intelligence. Given the inherent trade-off between efficiency and generality, we can understand this low efficiency as the Cost …

jp (code pays fourni par la source)

Reward-Agnostic Learning of Locomotion in Individualised Bodies Using Reward as an Evolutionary Stepping Stone

Solvi Arnold, Reiji Suzuki, Takaya Arita

Efficient learning is a defining feature of biological intelligence that remains difficult to replicate in artificial systems. We recently hypothesised that prior evolution of low-efficiency reward-driven learning can facilitate subsequent evolution of efficient reward-agnostic observation-driven learning for various task domains. However, this …

jp (code pays fourni par la source)

Agent-Based Models of Sexual Selection in Bird Vocalizations Using Generative Approaches

Takaya Arita, Reiji Suzuki

The current agent-based evolutionary models for animal communication rely on simplified signal representations that differ significantly from natural vocalizations. We propose a novel agent-based evolutionary model based on text-to-audio (TTA) models to generate realistic animal vocalizations, advancing from VAE-based real-valued genotypes to …

jp (code pays fourni par la source)