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 tremor in his hands made it impossible to knot the tie. He fumbled with the silk, the familiar pattern mocking him. Across the way, the splintered remains of Mrs. Gable’s ancient oak lay sprawled, half across her fence. He hadn't even heard it fall.

He'd been inside, engrossed in a particularly vapid online game. The rumble in the earth, the earth-shattering crunch - all missed. Now, the guilt constricted his chest like a too-tight vest.

He swallowed hard and looked out the window. Mrs. Gable was surveying the damage, her face a mask of weary acceptance. He wanted to help, to offer something, anything, but his feet felt like lead. He’d just stay inside, and pretend he hadn't noticed.

Emotion: ashamed

Cluster: Shame / Guilt
PC1 (Valence): -1.97 Negative
PC2 (Disposition): 1.06

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.376
ashamed0.344
ness0.327
逃0.297
😞0.295
Suppressed:
de-0.539
que-0.300
😍-0.288
(!)-0.269
🤩-0.242