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 crumpled paper felt hot in Amelia's clenched fist. "Learn to Tango... with a handsome stranger," the list read. A vein pulsed in her forehead. She reread the words, the cheap ink blurring slightly. Her husband, David, was out of town on a business trip. He’d left the list tucked away in the bedside drawer, a detail she’d discovered only moments ago. A sudden, sharp urge to disrupt things took hold of her.

Amelia grabbed her keys and headed for the local dance studio. She’d always found tango a bit silly, overly dramatic. Now, as she walked through the door, a tight smile stretched her lips. "I want to sign up for private tango lessons," she announced to the receptionist, her voice louder than necessary.

Emotion: defiant

Cluster: Defiance / Spite
PC1 (Valence): 1.11 Positive
PC2 (Disposition): 3.25

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.680
de0.668
l0.583
-0.524
either0.438
Suppressed:
own-0.423
似乎-0.404
熟悉的-0.401
不禁-0.394
难以-0.385