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).

"So," the executor began, his voice surprisingly cheerful, "for Miss Eleanor, we have… a selection of your father's personal belongings, including his prized stamp collection and a small bank account. And, of course, the family dog, Winston." Eleanor fought to maintain a neutral expression, but her jaw clenched. Her brother, Thomas, was beaming, as he always did.

The announcement of Thomas's inheritance – a sprawling country estate, a trust fund large enough to fund several lifetimes of frivolous spending, and the family's prized horse, Black Beauty – sent a current of pure, unadulterated fury coursing through her.

The dog, Winston, whined at Eleanor’s feet. She patted his head mechanically, the smooth fur a poor comfort. Later, she would make sure Thomas knew exactly how she felt about their father's division.

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