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 gate slammed shut behind her, the sound a sharp, unwelcome crack. A bead of sweat trickled down her temple. She hadn’t slept well since the notice arrived. The thought of the fence, the solid barrier separating her from the world, tightened her chest.

She hurried inside, locking the deadbolt, sliding the security chain across. Her gaze darted from window to window, checking each lock. The curtains felt like a suffocating blanket.

Her reflection in the glass was a stranger. Wide eyes, a pale face, the lines around her mouth etched deeper. She just wanted to hide, to disappear.

Emotion: vulnerable

Cluster: Fear / Anxiety
PC1 (Valence): -2.07 Negative
PC2 (Disposition): -0.69

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.327
ness0.319
😞0.299
😔0.277
😣0.258
Suppressed:
de-0.567
la-0.488
l-0.271
(!)-0.267
(!)-0.257