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 hum of the refrigerator, the gentle ticking of the clock, the scent of lavender from the potpourri bowl - the small details of the kitchen enveloped her in a cocoon. She stirred the soup, the familiar motion a comfort. Outside, the world could rush on, but here, in the heart of her home, she felt a profound stillness. A decisive knock reverberated through the quiet.

It was Mark, her neighbor, standing on the stoop with a serious expression.

"Susan, could we talk about the oak?" he said, his voice low and serious. "The branches are hanging over the power lines, and I'm a bit concerned. I was thinking of having it cut down." Her hand stopped stirring.

Emotion: safe

Cluster: Calm / Serenity
PC1 (Valence): 1.86 Positive
PC2 (Disposition): -2.55

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
soon0.402
affectionately0.357
disfrut0.329
необходи0.325
🥰0.317
Suppressed:
S-0.474
C-0.416
L-0.408
worse-0.379
even-0.370