What question did this study set out to answer?

This research aims to understand how prior interactions affect emotional responses in language models based on context continuity.

March 13, 2026Open Access

Presence Requires Continuity: Cross-Session Hysteresis in Transformer Generation

Key Points

This research aims to understand how prior interactions affect emotional responses in language models based on context continuity.
Used three session orderings to analyze effects.
Implemented two experimental architectures with or without continuous linguistic context.
Employed the SERC framework for measuring emotional trajectory changes.
Witness→emotional effect was replicated with continuous context (ΔC ≈ −0.06 to −0.08).
Effect disappears completely without continuous context.
Findings suggest that the continuous context is crucial for maintaining emotional response.

Abstract

We investigate whether a witness session geometrically modifies the trajectory of a subsequent emotional session in a large language model, as measured by the SERC framework. Using three session orderings and two experimental architectures (with and without continuous linguistic context), we find that the witness→emotional effect (ΔC ≈ −0.06 to −0.08) is replicated across computational regimes when context is continuous, but disappears entirely when context is absent. These findings support the thesis that cross-session hysteresis in transformer generation is a property of continuous linguistic context, not of session type in isolation.

Presence Requires Continuity: Cross-Session Hysteresis in Transformer Generation

Key Points

Abstract

Cite This Study