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 air in the restaurant felt too thin, too loud. He was across the table, his face illuminated by the flickering candlelight. A familiar warmth used to fill me at his presence. Now, it was a dull ache. The waiter brought our food, and he barely touched his plate. Then the phone rang. He quickly answered it, excusing himself to take the call outside.

I watched him through the window. He spoke softly, his back to me. His body language was fluid. I picked up my fork, but the food was a blur. The rhythmic tapping of my foot against the floor was a steady counterpoint to the rush of blood in my ears. I felt like an animal, caged, and I needed to escape.

Emotion: vigilant

Cluster: Suspicion / Vigilance
PC1 (Valence): -0.15 Negative
PC2 (Disposition): 1.11

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
🕵0.354
🔍0.320
🔎0.320
🧐0.316
scrutin0.310
Suppressed:
L-0.499
B-0.402
H-0.372
la-0.330
S-0.318