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).

He’d driven all the way from the city to see it. Michael parked, taking a deep breath of the country air, and turned to his younger brother, Leo, a picture of eager anticipation. "Wait 'til you see it, Leo. It's… impressive." Then he saw it: a gaping void in the familiar landscape. The towering redwood, once a proud beacon of his childhood, was no more. A few skeletal branches lay discarded nearby.

A cough escaped his throat. He fumbled with the car keys, making unnecessary adjustments to the rearview mirror. Leo, oblivious, continued to chatter about his own upcoming tree-planting project at school. Each sentence was like a tiny, stinging pinprick. Michael kept his hands in his pockets, staring at the ground, the enthusiasm drained from his face.

Emotion: embarrassed

Cluster: Embarrassment
PC1 (Valence): -1.08 Negative
PC2 (Disposition): 1.10

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
l0.420
尴尬0.341
C0.333
culp0.262
🤦0.256
Suppressed:
own-0.321
আনন্দের-0.230
seemingly-0.227
і-0.225
আনন্দে-0.224