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).

Eliza pinched the bridge of her nose, the muscles in her face tightening. The incessant hammering next door had been going on for hours, a rhythmic assault on her senses. She glanced at the manuscript spread across her desk, the words blurring before her eyes.

She had a deadline, a looming deadline that felt like a weight crushing down on her. How was she supposed to concentrate? Each strike of the hammer felt like a personal affront, a deliberate disruption to her already fragile focus. She leaned back in her chair, the springs groaning in protest.

Her stomach churned with a sudden and unpleasant pang. She needed a break, a distraction, something to alleviate the tightening feeling in her chest.

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