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 drive back from the job fair was agonizing. Liam was in the passenger seat, chattering about the brilliant presentation he’d given, the favorable impression he'd made.

I gripped the steering wheel so tightly my knuckles ached. The radio blared a cheerful pop song, adding to the feeling of pressure in my temples. Each lane change felt like a struggle, each red light a test of my dwindling patience.

He was oblivious, of course. Lost in his own world of self-approval. I took a deep breath, trying to tamp down the rising tide of… displeasure.

"That was a great day!" he said, the pure joy in his voice grating on my nerves. I remained silent.

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