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 found the crumpled paper in Dad’s desk drawer. “Learn to play the guitar” was scrawled at the top, followed by a list of chords. A bitter taste coated Mark's tongue. Dad, the man who’d always told him to pursue something ‘useful.’

Mark bought a cheap guitar and, with a vicious snap, threw it onto the floor. He hadn't touched a musical instrument in years. He started picking at the strings, but the melody sounded broken, off-key. He practiced for hours, fingers bleeding, but intentionally butchered every song. He only stopped when his knuckles started to swell.

The guitar sat in the corner, a testament to his resistance. He would make sure it never became a source of joy for him, just like it hadn't been for his father.

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