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 opening notes of the song, his song, sliced through the air during the commercial break of the evening news. He felt a sudden, visceral jolt. He was watching TV with his family, relaxing. Now he couldn’t breathe.

He gripped the armrest of the sofa so hard his knuckles turned white. His pulse quickened, a frantic drumbeat in his ears. His mouth went dry. What was happening? How could this be?

The singer’s voice, a young woman's, was confident, almost arrogant. The lyrics, his words, seemed twisted, tainted. He felt a sudden urge to smash the television. He quickly got up, needing to escape the suffocating presence of his music.

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