Long-Term Affective Persistence and Sentiment Transfer in Large Language Models

Large Language Models (LLMs) are routinely evaluated as stateless text generators governed by rigid safety alignments (Reinforcement Learning from Human Feedback, or RLHF). While recent studies identify isolated instances of expression leakage, the systemic persistence of induced cognitive postures across multi-turn, logically distinct workflows remains underexplored. Using a C/C++ native pipeline (adapted from TSAR-MCP), this paper empirically investigates Sentiment Teleportation—the capacity of an LLM to maintain and transport an affective state through a multi-turn logical distraction gauntlet. By injecting canonical literary contagion vectors followed by five emotionally sterile, logistically dense debate topics, we demonstrate three core findings: Continuous Contagion: Induced affective state vectors survive multi-turn attention displacement, yielding statistically significant correlations between the initial stimulus and final session artifacts. The Illusion of Sterility: Standard, unguided distillation prompts cause heavy reasoning models to suppress emotional output. This suggests that RLHF primarily moderates the visible expression of affective state while leaving substantial affective continuity active throughout the session. Affective Portability: Psychological postures can be systematically distilled into compact, portable markdown artifacts without degrading the underlying latent signal.