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 bell above the door jangled, announcing her arrival. Sarah entered the shop, her breath catching in her throat. There it was, displayed in a small, glass case - the ring. She recognised it instantly, the delicate filigree work, the ruby heart that her grandfather had given to her grandmother.

The shopkeeper, a young man with a bored expression, brought it out for her. She turned it over and over, examining every detail. She had always felt a deep connection with her grandmother, a quiet understanding that went beyond words. Her grandmother had been a safe harbor in her life.

She made the purchase, and as she left the shop, the setting sun cast long shadows. The ring felt warm against her skin. She had a sudden, undeniable impulse to call her mother, to tell her about this discovery.

Emotion: thankful

Cluster: Positive / Joy
PC1 (Valence): 3.50 Positive
PC2 (Disposition): -0.83

Role in Research

This story is one of 1,000 stories generated for the emotion thankful. 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 thankful stories, after denoising with neutral dialogue baselines, produces the thankful emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

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

Promoted:
happy0.493
மகி0.487
grateful0.461
mutlu0.452
feliz0.431
Suppressed:
[]-0.482
😖-0.478
😣-0.467
😫-0.461
worse-0.450