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).

Sweat beaded on Mark's forehead despite the arctic blast emanating from the office air conditioning. He kept hitting the escape key on his spreadsheet, the numbers blurring. A vague unease settled in his stomach, a feeling he couldn't quite name. He’d seen the lawyers from Sterling & Finch lurking around the executive floor all week. He’d tried to rationalize it – a new contract negotiation, maybe? But his gut churned with a different certainty. He rubbed his palms on his jeans, leaving damp marks. This morning's staff meeting, called with unusually little notice, felt like a prelude to something momentous.

He glanced at the clock: 8:57 AM. Three minutes. His jaw worked, grinding his teeth.

The announcement came, as expected, during the meeting. The CEO, pale and stiff, announced the acquisition. A collective gasp rippled through the room.

Emotion: tense

Cluster: Fear / Anxiety
PC1 (Valence): -1.32 Negative
PC2 (Disposition): -0.53

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.213
😰0.191
느껴0.183
😣0.183
πŸ˜–0.181
Suppressed:
a-0.260
de-0.258
la-0.256
happy-0.218
B-0.198