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 photograph, a faded black and white print, was clutched tightly in Chloe’s hand. A smiling woman, her mother’s double, posed with a man she didn't recognise. This was her "aunt", apparently. Chloe's world felt askew. Her mother’s silence about her past, her father’s careless betrayal – it all converged in this single, damning photograph.

Chloe crumpled the picture into a ball and threw it across the room. It landed near the hearth.

She walked into the kitchen and poured herself a glass of wine, her hand shaking as she did so. Every piece of family history she thought she knew was a lie. Chloe wanted to scream. Instead, she stared out the window, at the gray, indifferent sky.

Emotion: scornful

Cluster: Anger / Hostility
PC1 (Valence): -0.76 Negative
PC2 (Disposition): 2.07

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
l0.406
C0.391
S0.276
I0.268
愤0.243
Suppressed:
own-0.396
unexpectedly-0.288
Suddenly-0.260
soon-0.256
serendip-0.256