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 avoided eye contact with Mrs. Henderson at the mailbox. The womanโ€™s unasked questions, the sympathetic glances, only amplified his discomfort. He hurried back inside, the door slamming shut with a finality that felt both liberating and isolating. The hammering next door had started again.

He went directly to his room and sat on the edge of the bed. The weight of unspoken grief pressed down on him. He felt hollowed out, a shell of his former self, and the renovations next door seemed to be a constant, painful commentary on his own unfinished project of life.

Emotion: sorry

Cluster: Sadness / Despair
PC1 (Valence): -2.02 Negative
PC2 (Disposition): -0.36

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.604
L0.549
๐Ÿ˜ž0.382
๐Ÿ˜”0.364
ness0.324
Suppressed:
de-0.675
la-0.536
(!)-0.323
!-0.298
l-0.288