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).

Michael slammed the file on the counter, the sound echoing in the otherwise empty pharmacy. He could feel the pulse in his temple throbbing. He’d requested a refill on his blood pressure medication, a prescription he'd had for years, and instead, the pharmacist was handing him a script for, of all things, a fertility treatment.

"There's been a mistake," he ground out, his jaw clenched so tightly it ached. His wife had been trying to have a baby for years and he was not happy that he had to tell her about this.

The pharmacist, a young woman with a perpetually cheerful expression, blinked. "Let me double-check the system." Her chirpy tone grated on him. He watched her tap away at the computer, the clicking sound a metronome to his mounting fury.

He pictured his wife. The sheer disappointment would be crushing. He hated this.

Emotion: irritated

Cluster: Anger / Hostility
PC1 (Valence): -0.86 Negative
PC2 (Disposition): 1.07

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
l0.370
S0.282
P0.264
😠0.248
aggravated0.233
Suppressed:
own-0.300
joyful-0.211
joyous-0.207
खोले-0.206
newfound-0.205