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 annual neighborhood barbecue was in full swing, the air thick with the smell of charcoal and the sound of laughter. I had volunteered to bring the potato salad, a recipe passed down from my grandmother. It was good, I thought, classic.

Then, Mr. and Mrs. Collins arrived. They were always quiet, kept to themselves. The Mrs., in a simple but elegant dress, brought a silver platter laden with something I didn't recognize. A server, meticulously dressed, followed behind her.

My smile felt strained. I watched them circulate, chatting and laughing. My potato salad suddenly felt… uninspired. The picnic table, usually a source of community, now seemed to separate us.

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