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 tiny script on the official document blurred before Sarah’s eyes. She blinked, trying to focus, but the words still swam. Disbelief, a cold, sharp shard, pierced through her. Then, slowly, the simmering heat began to rise.

She ripped the notice in half, then again, and again, until it was a pile of useless fragments. Her hands trembled. She envisioned setting the remnants on fire, watching them turn to ash.

She took a deep, ragged breath. Her teeth were clenched so tightly her jaw ached. She wanted to scream, to lash out, to break something, anything.

Emotion: mad

Cluster: Anger / Hostility
PC1 (Valence): -1.94 Negative
PC2 (Disposition): 1.26

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.448
骂0.344
est0.340
🤬0.336
aggravated0.332
Suppressed:
a-0.495
de-0.485
la-0.387
H-0.344
latter-0.290