Towards LifeSpan Cognitive Systems

TL;DR

This paper proposes a conceptual framework for LifeSpan Cognitive Systems (LSCS) that can continuously learn, retain, and recall experiences in complex environments, focusing on large language models and integrating multiple storage technologies.

Contribution

It introduces a new conceptual model for LSCS that combines four classes of experience storage technologies to enable continuous, high-frequency learning and recall in language models.

Findings

Identifies two major challenges: abstraction and long-term retention.

Classifies existing technologies based on storage complexity.

Proposes an integrated instantiation for LSCS with core processes.

Abstract

Building a human-like system that continuously interacts with complex environments -- whether simulated digital worlds or human society -- presents several key challenges. Central to this is enabling continuous, high-frequency interactions, where the interactions are termed experiences. We refer to this envisioned system as the LifeSpan Cognitive System (LSCS). A critical feature of LSCS is its ability to engage in incremental and rapid updates while retaining and accurately recalling past experiences. In this paper we focus on the domain of Large Language Models (LLMs), where we identify two major challenges: (1) Abstraction and Experience Merging, and (2) Long-term Retention with Accurate Recall. These properties are essential for storing new experiences, organizing past experiences, and responding to the environment in ways that leverage relevant historical data. Unlike language models with continual learning, which typically rely on large corpora for fine-tuning and focus on improving performance within specific domains or tasks, LSCS must rapidly and incrementally update with new information from its environment at a high frequency. Existing technologies with the potential of solving the above two major challenges can be classified into four classes based on a conceptual metric called Storage Complexity, which measures the relative space required to store past experiences. Each of these four classes of technologies has its own strengths and limitations while we argue none of them alone can achieve LSCS alone. To this end, we propose a potential instantiation for LSCS that can integrate all four classes of technologies. The new instantiation, serving as a conjecture, operates through two core processes: Absorbing Experiences and Generating Responses.