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 harsh fluorescent lights of the storage unit cast long, unsettling shadows. David watched his sister, Olivia, already reaching for a key to a new storage space. The lawyer announced, "Olivia receives the family's classic car collection, including the vintage Ferrari." Olivia squealed with glee, her eyes sparkling. David felt a strange emptiness settle within him.

He steeled himself to hear his portion. "David, you inherit the family's collection of vintage stamps." A coldness spread through him, creeping up from his toes. Stamps? He swallowed hard, his throat constricted. He could not bring himself to meet Olivia’s gleeful eyes. He felt a powerful urge to run, to escape the musty, confined space. He took a shaky step back.

Emotion: alarmed

Cluster: Fear / Anxiety
PC1 (Valence): -2.07 Negative
PC2 (Disposition): -0.14

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😰0.304
😨0.272
worse0.269
😣0.261
не0.254
Suppressed:
de-0.323
enthusi-0.258
happy-0.235
delight-0.225
ية-0.217