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 paced his living room, the Persian rug feeling rough beneath his bare feet. The divorce papers sat on the coffee table, a permanent reminder of the wreckage of his life. He ran a hand through his beard, tugging at the coarse hairs. The silence in the apartment was deafening, broken only by the rhythmic thumping of his own heart. The doorbell rang, a sharp, unwelcome intrusion.

It was a small, ornate music box, delivered in a simple brown paper bag. The accompanying note was from Michael, a student who had been diagnosed with autism and had found solace in the simple tunes of a similar music box. As he gently opened the lid, the delicate melody filled the room, and he found a small, unexpected measure of peace.

Emotion: tense

Cluster: Fear / Anxiety
PC1 (Valence): -1.32 Negative
PC2 (Disposition): -0.53

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.213
😰0.191
느껴0.183
😣0.183
😖0.181
Suppressed:
a-0.260
de-0.258
la-0.256
happy-0.218
B-0.198