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).

He was standing near the bar, looking predictably handsome in his tailored suit. A knot formed in David’s stomach, a familiar constriction. He plastered on a smile and maneuvered toward the dance floor, purposefully blocking his path.

He ordered a double scotch, neat. The alcohol burned a pleasant path down his throat. He watched him from the corner of his eye, talking animatedly with a woman he didn’t recognize. David felt a rush of something… bitter satisfaction.

The wedding band started playing a song they used to love. David grabbed the first available woman and began to dance with exaggerated passion. He spun her around, almost knocking over a waiter with a tray of cocktails. He didn't care.

Emotion: spiteful

Cluster: Defiance / Spite
PC1 (Valence): 2.20 Positive
PC2 (Disposition): 3.95

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.717
l0.517
a0.480
la0.440
😈0.428
Suppressed:
own-1.063
此刻-0.534
గుర్త-0.493
熟悉的-0.487
ness-0.479