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).

"Absolutely unbelievable!" David slammed the book down on his desk, the action more vehement than he intended. The audacity of it! Someone, a total stranger, had meticulously documented his life, his career, his relationships, all without so much as a by-your-leave. He couldn’t shake this odd energy buzzing in his veins.

He had to do something. He paced his office, a tight knot forming in his stomach. He had an urge to destroy something. Then, an idea sparked, a sudden, bright thought. He would fight back.

He grabbed his phone, a grin spreading across his face. He’d spend all night on the phone with the lawyers. He dialed, the dial tone a sharp, insistent sound in the quiet office. He looked forward to every second of it. This was war, and he was ready.

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