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).

Rain lashed against the window of his studio. He slammed the door, the sound echoing in the high-ceilinged room. They’d done it. Smeared his painstakingly rendered ink wash across someone’s *skin*. He felt a tremor in his hand, a twitch that made his fingers curl inward.

He grabbed a bottle of scotch and poured a generous amount. He’d poured himself a drink before, sure, but never with this bitter taste in his mouth.

He pictured the smug look on that tattoo artist’s face. He could feel a familiar churning in his stomach, the sensation of being utterly disregarded.

He stalked towards his easel, kicking a stray paint tube out of his path. The canvas, half-finished, seemed to mock him with its vulnerability. He took a long swallow of the whiskey and glared at the painting, contemplating tearing it down.

Emotion: disdainful

Cluster: Anger / Hostility
PC1 (Valence): -0.31 Negative
PC2 (Disposition): 1.79

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.419
l0.404
S0.387
T0.282
I0.257
Suppressed:
own-0.335
unexpectedly-0.285
Suddenly-0.246
previously-0.245
until-0.242