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’d been cleaning out his late father’s desk. Arthur, a man of routine and quiet contemplation, hadn’t expected to unearth anything truly shocking. Then he found it: a letter, addressed to a woman he’d never heard of, dated from the summer before his parents married. His breath hitched.

He reread the lines, his eyes widening with each sentence. The confession of a secret, passionate affair. He felt a dizzying sensation, like the world had tilted on its axis. He slowly sat, the chair groaning under the sudden weight. His father, the stoic, reserved man he’d known, had held a hidden life? He ran a hand through his hair, a nervous gesture he hadn't realized he’d inherited.

Emotion: astonished

Cluster: Surprise / Confusion
PC1 (Valence): -0.84 Negative
PC2 (Disposition): -0.24

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
que0.448
de0.350
la0.323
l0.307
incred0.266
Suppressed:
own-0.289
cheerfully-0.231
enjoyed-0.224
开心-0.221
ี่ยว-0.219