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).

John scrubbed at his forehead, the sweat stinging his eyes. He found the list tucked between the mattress and the box spring. It was in Emily’s perfect, looping handwriting. *Ride a Harley Davidson across Route 66.* He felt a pulse hammering in his temple.

He threw the list onto the floor. Dust motes danced in the afternoon sunbeams, momentarily catching his eye. *Write a novel*. She’d always scoffed at his attempts to start a blog.

He stalked across the room, kicking aside a discarded shoe. Why? Why hadn't she ever mentioned any of this? The air felt thick, heavy, like a suffocating blanket. He balled his hands into fists, nails digging into his palms.

Emotion: enraged

Cluster: Anger / Hostility
PC1 (Valence): -2.14 Negative
PC2 (Disposition): 1.33

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.555
est0.370
aggravated0.362
骂0.359
🤬0.345
Suppressed:
a-0.468
de-0.358
la-0.312
latter-0.287
☺️-0.285