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 beam of her flashlight danced across the cluttered floorboards. Her fingers drummed a frantic rhythm against her thigh. She'd been putting this off for weeks. Now, here she was. The air hung still and lifeless.

A child’s rocking horse, chipped and faded, stood in the corner. She averted her gaze. It felt like someone was watching her. She had to take a deep, steadying breath.

She saw a collection of old toys, and she flinched at the sight. They looked so out of place, abandoned. She reached for a tarnished music box, her hand trembling. The anticipation was almost unbearable.

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
😖0.181
Suppressed:
a-0.260
de-0.258
la-0.256
happy-0.218
B-0.198