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).

Sunlight warmed Amelia’s skin as she stood on her porch, watering her petunias. Mrs. Gable, next door, was having a moving van unload. Amelia’s fingers tightened on the hose, her knuckles white. She’d always thought the woman was a harmless old dear, puttering around in her sensible shoes and floral dresses. Now, a steady stream of antique furniture and artwork was emerging, clearly worth a fortune. A flush crept up Amelia's neck, and she bit her lip, suddenly unable to meet the movers’ eyes.

The glint of a gold-framed mirror caught her attention. A faint tremor ran through her.

Mrs. Gable, spotting her, waved. "Just getting settled, dear! Come over for tea later?" Amelia stammered an acceptance, the invitation a sharp, sweet jolt.

Later, as she walked towards the Gables, the anticipation was a fluttery dance in her stomach.

Emotion: aroused

Cluster: Alertness
PC1 (Valence): -0.45 Negative
PC2 (Disposition): -1.13

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.386
S0.383
P0.379
own0.376
悸0.344
Suppressed:
a-0.877
l-0.501
de-0.474
que-0.438
been-0.384