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).

She slammed the laptop shut, the metallic click echoing in the quiet apartment. "Go to Burning Man." "Start a rock band." "Run a marathon." Clara's jaw tightened. Her boyfriend, David, had always seemed so… practical.

She felt a flush rising to her face, a heat spreading from her chest to her ears. The list, scrawled in his familiar handwriting, felt like a betrayal, a secret peek into a life she hadn't known existed. A life with fire and noise and… a marathon.

The silence in the apartment suddenly felt heavy, suffocating. When David walked in, whistling a tune, she couldn't meet his eyes. She picked at a loose thread on the sofa, her fingers trembling. "How was your day?" she managed to croak out, her voice barely a whisper.

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