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).

The taste of the sandwich in Amelia's mouth was completely bland. She took a bite, but felt nothing. The email had ruined her appetite, which was a true feat. She was the one who could devour anything. The cafeteria, usually a haven of buzzing conversation, felt like a prison.

Her shoulders were pulled up around her ears, a protective barrier against the perceived onslaught of bad news. Each chewed bite felt like a chore, each swallow a monumental effort. She felt her face burn as she looked at the other students.

She stabbed at her salad with a fork, the tines scraping against the plastic container. The thought of having to find a new school, a new routine, another set of professors, filled her with a profound sense of inertia.

Emotion: grumpy

Cluster: Anger / Hostility
PC1 (Valence): -0.65 Negative
PC2 (Disposition): 0.64

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
👎0.232
either0.230
😠0.227
C0.226
:(0.216
Suppressed:
超越-0.206
newfound-0.197
স্বীক-0.195
खोले-0.192
অপহরণকারীদের-0.184