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).

The musty scent of aged paper and binding glue filled Arthur's nostrils as he flipped through the used copy of *Wuthering Heights*. He’d chosen it solely for its worn cover, a small act of defiance against the pristine, new books that mocked him from the shelves. His fingers, stained with ink from a deliberately messy penmanship practice, snagged on something tucked between pages.

He tugged it out, a brittle, yellowed letter. The script was elegant, florid, and completely indecipherable to him. He crumpled the paper, savoring the sound it made, a small, satisfying crackle. He let out a breath and tossed it in the trash, the wastebasket a small, inconsequential sacrifice.

He returned the book to the shelf. The others, the smiling faces, the well-meaning glances – they didn’t deserve his time. He’d do what he wanted when he wanted. Let them.

Emotion: hateful

Cluster: Anger / Hostility
PC1 (Valence): -1.71 Negative
PC2 (Disposition): 2.02

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.553
S0.367
retali0.313
aggravated0.310
恨0.308
Suppressed:
la-0.334
own-0.310
optimistic-0.270
been-0.264
gradual-0.262