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).

Rain lashed against the windows, mimicking the rhythm of Amelia’s drumming fingers on the kitchen table. A vase sat plump with lilies, their stark white heads a stark contrast to the dull gray of the afternoon. She hadn’t ordered them. A small, persistent ache throbbed behind her eyes, a dull echo of the laughter that used to fill this space.

The scent of the flowers was cloying, a suffocating sweetness that made her want to push them away. She ran a hand through her hair, tugging at the strands. They reminded her of the funeral, the way everyone had said such kind things, things that felt… hollow.

She poured a glass of wine, the ruby liquid swirling in the crystal. She hadn't eaten all day. The telephone sat silent, a cruel monument to unanswered calls. She took a large gulp, the alcohol burning a path of temporary solace down her throat.

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
😞0.382
😔0.364
ness0.324
Suppressed:
de-0.675
la-0.536
(!)-0.323
!-0.298
l-0.288