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 envelope felt strangely heavy in her hands, as if it contained not just paper, but a weight of unspoken desires. Amelia sat on the edge of her father's bed, the scent of his aftershave still lingering in the air. The list was tucked inside, and a small, dried flower fell out, a pressed forget-me-not.

She started reading. "See the Northern Lights." A tear traced a path down her cheek. That was her Dad! The very man who always said camping was "too much like work." Why hadn’t he told her? She scanned the next items: "Learn Spanish," "Write a novel," "Fall in love again." Her chest felt tight.

Amelia slowly got up, her legs wobbly. She looked around the room, each object a painful reminder. The empty space felt wider, echoing the emptiness inside her. Where had she been when her father wanted to see the Aurora Borealis?

Emotion: shaken

Cluster: Fear / Anxiety
PC1 (Valence): -2.72 Negative
PC2 (Disposition): -0.86

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.351
unable0.269
focus0.266
😞0.258
다시0.257
Suppressed:
de-0.769
la-0.579
a-0.375
happy-0.292
"-0.290