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 gallery's opening night for his newest exhibition was supposed to be the culmination of years of work. Now, it was a blur of forced smiles and strained conversation. The air hung thick with the scent of expensive perfume and unspoken tension. He felt a persistent throb behind his eyes. He saw the tattoo on the back of his agent's hand: a miniature copy of "Whispers in the Wind."

The casual arrogance of the act felt like a physical blow. He excused himself, escaping to the balcony. The city lights twinkled below, but he saw no beauty in them. The once-bright promise of his work now felt tarnished, defiled. He felt as though the wind had been knocked out of him.

Emotion: miserable

Cluster: Sadness / Despair
PC1 (Valence): -2.19 Negative
PC2 (Disposition): -0.87

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😞0.369
无法0.321
無法0.270
💔0.265
😔0.262
Suppressed:
de-0.490
la-0.329
B-0.300
l-0.292
/-0.224