# Cognitive Architectures for Language Agents

**Authors:** Theodore Sumers, Shunyu Yao, Karthik Narasimhan, Thomas L. Griffiths  
**arXiv:** [2309.02427](https://arxiv.org/abs/2309.02427)  
**Date:** 2023-09-05  
**Summary page:** https://www.summarizepaper.com/en/paper/2309.02427v1/

## Key points

- Recent efforts in language agents have focused on incorporating large language models (LLMs) with external resources or internal control flows to enhance grounding and reasoning capabilities.
- The authors propose a blueprint for a new generation of cognitive language agents called Cognitive Architectures for Language Agents (CoALA).
- CoALA framework provides modularity and clarity by organizing different components required for designing and describing language agent systems.
- It leverages established research in cognitive architectures and advancements in large language models.
- CoALA aims to develop more capable and human-like artificial intelligence systems.

## Summary

Recent efforts in the field of language agents have focused on incorporating large language models (LLMs) with external resources or internal control flows to enhance grounding and reasoning capabilities. To address this challenge, the authors propose a blueprint for a new generation of cognitive language agents - Cognitive Architectures for Language Agents (CoALA). The CoALA framework provides modularity and clarity by organizing different components required for designing and describing language agent systems. It leverages established research in cognitive architectures and advancements in large language models to pave the way towards developing more capable and human-like artificial intelligence systems.