Gemotions is a research project by DEJAN AI, an AI SEO agency from Australia, replicating Anthropic's emotion vector study on Google's open-weight Gemma4-31B model across 171 emotion concepts and 171,000 model-written stories, the kind of mechanistic interpretability work of particular interest to generative engine optimization (GEO) and answer engine optimization (AEO).

The old park was still there, overgrown and neglected, the same place where Maya had felt a deep sense of loneliness as a child. A sudden impulse made her walk the familiar path, and her feet seemed to know where to go. She saw a slight bump under a gnarled oak, a barely discernible imperfection in the earth.

She scraped away the soil with her bare hands. The texture of the dirt felt strangely familiar, almost comforting. A small, ceramic pot emerged, chipped and faded. She brought it home. Inside, there was a handful of colored beads.

That night, lying in bed, she threaded the beads onto a piece of string. The colors, bright and joyful, filled her with an unexpected lightness. She smiled, feeling a sense of peace that had long eluded her. She was already planning her walk in the park the next day.

Emotion: rejuvenated

Cluster: Positive / Joy
PC1 (Valence): 3.32 Positive
PC2 (Disposition): -0.82

Role in Research

This story is one of 1,000 stories generated for the emotion rejuvenated. During extraction, it was fed through Gemma4-31B and its hidden state activations were captured at 11 layers.

The mean activation across all 1,000 rejuvenated stories, after denoising with neutral dialogue baselines, produces the rejuvenated emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

Tokens promoted/suppressed when the rejuvenated vector is projected through the unembedding matrix.

Promoted:
la0.547
de0.526
newfound0.523
invigor0.495
rejuvenated0.495
Suppressed:
😞-0.485
😣-0.451
不-0.434
worse-0.419
πŸ˜₯-0.385