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 sun beat down on the patio, making the air thick and heavy. Andrew ran a hand through his hair, already feeling sticky. He'd agreed to meet Alex for a celebratory brunch. The celebrations, however, seemed premature.

Alex had been boasting about his new consulting job for months. But Andrew had caught a glimpse of Alex’s work laptop screen the other day, and the spreadsheets seemed more… administrative than strategic.

“So, the big consulting gig… everything going well?” Andrew asked, forcing a smile.

Alex took a long pull of his mimosa, his face unreadable. "Couldn't be better. Challenging, sure, but rewarding." He tapped his fingers on the table, a nervous rhythm.

Emotion: tense

Cluster: Fear / Anxiety
PC1 (Valence): -1.32 Negative
PC2 (Disposition): -0.53

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.213
😰0.191
느껴0.183
😣0.183
😖0.181
Suppressed:
a-0.260
de-0.258
la-0.256
happy-0.218
B-0.198