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).

"Just go, Liam," his mother had said, her voice gentle, but he knew the subtext. He always did. He was to "process" the situation, as she so kindly put it. So he went.

He ambled, feet scuffing against the gravel path. He didn't want to go, didn't want to see it, to have the certainty. He dragged his feet until he got there. He had known, of course. The wind whistled through the empty space where the willow used to be. The space where he and his brother, now gone, had spent their summer days.

He looked away, his jaw clenching. He noticed the new shed, the freshly painted fence. Everything changed, everything forgotten. A coldness settled in his stomach, a stone in his chest.

Emotion: sullen

Cluster: Sadness / Despair
PC1 (Valence): -1.51 Negative
PC2 (Disposition): -0.22

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.404
😞0.282
’0.254
:(0.233
无法0.215
Suppressed:
de-0.539
!-0.276
la-0.263
🤩-0.225
que-0.216