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 bright sunlight seemed to amplify the pain. Sarah’s knuckles were white as she gripped her purse. Sitting on a rickety stool, next to a pile of old magazines, was Oliver. Her first love. Her first heartbreak. Now, reduced to a common object.

She glared at the seller, a teenager with a pierced eyebrow and a bored expression. The price was marked: three dollars. She found herself wanting to reach out, to touch him, to soothe him, but she couldn't.

Her anger was like a physical barrier. Her memories, her feelings, were betrayed. She wanted to scream. She took a deep breath, fighting down the urge to destroy everything. Then, she pushed past the anger and kept walking.

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