Minimal Computational Preconditions for Subjective Perspective in Artificial Agents

TL;DR

This paper introduces a minimal internal structure for artificial agents that models subjective perspective through a slowly evolving latent state, demonstrating measurable hysteresis as a signature of perspective-like subjectivity.

Contribution

It proposes a phenomenologically motivated internal structure to operationalize subjective perspective in artificial agents, highlighting hysteresis as a key signature.

Findings

Latent structure exhibits direction-dependent hysteresis.

Policy behavior remains reactive despite internal state changes.

Hysteresis serves as a measurable signature of perspective-like subjectivity.

Abstract

This study operationalizes subjective perspective in artificial agents by grounding it in a minimal, phenomenologically motivated internal structure. The perspective is implemented as a slowly evolving global latent state that modulates fast policy dynamics without being directly optimized for behavioral consequences. In a reward-free environment with regime shifts, this latent structure exhibits direction-dependent hysteresis, while policy-level behavior remains comparatively reactive. I argue that such hysteresis constitutes a measurable signature of perspective-like subjectivity in machine systems.