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).

He’d buried the past, or so he thought. Now, this. A belated mea culpa, delivered after a lifetime of silence. The sight of the envelope, elegant and understated, sent a current through his limbs. Robert's hands clenched, crumpling the letter in his grip, then quickly smoothing it out.

He felt a deep and visceral need to *do* something, anything. He strode to the piano, slamming the lid open with more force than necessary. His fingers hovered above the keys, his muscles tense. He began to play, a furious, chaotic melody that mirrored the turmoil within him. The music swelled, filling the room with a chaotic joy.

Emotion: stimulated

Cluster: Alertness
PC1 (Valence): 0.38 Positive
PC2 (Disposition): -0.18

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
own0.317
T0.312
intensity0.280
S0.271
inesis0.265
Suppressed:
a-0.640
l-0.434
la-0.392
que-0.378
fact-0.354