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).

From his living room window, Robert watched his neighbor, Ms. Gable, climb into a chauffeured SUV. He’d always known her as the woman who loved her cats and always had a kind word. A wave of desolation washed over him.

He replayed their shared moments in his mind: the casual conversations about the weather, the shared frustration about noisy leaf blowers. He’d trusted her, confided in her. The knowledge of her wealth created a strange vacuum in the space between them.

He found himself clenching his fists. The everyday annoyances, the mundane struggles – had they been a performance on her part? He felt foolish, naive, like he’d been living in a carefully constructed fantasy.

Emotion: hurt

Cluster: Fear / Anxiety
PC1 (Valence): -2.46 Negative
PC2 (Disposition): -0.42

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.412
L0.396
ness0.338
😞0.328
😢0.290
Suppressed:
de-0.643
la-0.402
!-0.259
(!)-0.248
/-0.244