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 aroma of cinnamon and stale coffee hung in the air of the old diner. Claire leaned against the counter, reading the letter she'd just found tucked in a cookbook. It had been written to a college friend, a confession of her secret feelings for him. The words, dripping with youthful angst, made her cringe.

She felt a cold wash over her, a deep chill that ran from her toes to the top of her head. She tucked the letter back into the cookbook, her fingers trembling. A waiter placed a cup of coffee in front of her, and she fumbled for a sugar packet. She was aware of every awkward angle of her body, every imperfection.

Emotion: humiliated

Cluster: Embarrassment
PC1 (Valence): -1.79 Negative
PC2 (Disposition): 0.93

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.370
l0.329
尴尬0.322
ness0.303
because0.298
Suppressed:
de-0.369
own-0.254
😍-0.253
🤩-0.245
cautiously-0.228