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 interviewer, that smirking little twit, was the first target. He'd been dispatched with a single, cutting remark about the biographical inaccuracies. Now, she was on her way to the publisher. Her stomach churned. The elevator felt impossibly slow, each floor a fresh torment. She imagined the smarmy smiles, the patronizing reassurances.

She walked into the building, her heels clicking a sharp rhythm against the marble floor. She met the receptionist’s gaze and felt a tremor in her hands. The air felt thin, and the scent of expensive perfume seemed to suffocate her. This was wrong, this intrusion. It was a violation.

Emotion: hostile

Cluster: Anger / Hostility
PC1 (Valence): -1.67 Negative
PC2 (Disposition): 1.56

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.592
S0.387
骂0.324
est0.301
aggravated0.295
Suppressed:
a-0.457
de-0.375
H-0.289
🤩-0.280
la-0.273