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 lawyer’s words echoed in Elara’s ears, each syllable a cruel hammer blow against her chest. “To Thomas, the family estate and all its holdings. To Elara… a collection of antique buttons.” She stared at her brother across the mahogany table. Thomas was already sketching plans for renovations, a smug look plastered across his face. A fire raged in her gut, twisting her stomach into knots. She abruptly pushed back from the table, the chair scraping sharply against the polished floor.

Outside, the autumn wind whipped around her, mirroring the turmoil within. Leaves, the color of dying embers, swirled around her ankles. She walked, head bowed, the weight of the injustice pressing down on her shoulders. Her hands clenched into fists, the nails digging into her palms. She would have to find a way to make sense of this.

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