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 attic air tasted of dust and forgotten things. Clara tugged at the string of the old lamp, its bare bulb cutting a harsh swathe through the gloom. She was searching for her grandmother’s wedding dress, a request that felt more like an obligation than a pleasure. A cardboard box, labeled “Dad – Personal,” caught her eye. She lifted the lid, revealing a jumble of yellowed papers.

Her breath hitched as she unfolded a sheet bearing her father’s elegant handwriting. It was a draft letter of resignation, addressed to the headmaster of his old boarding school. Dated 1982. The carefully crafted prose spoke of “irreconcilable differences” and a “lack of faith in the institution’s principles.” Her chest squeezed.

She ran a hand through her hair, a nervous gesture she'd inherited from him. He had always painted himself as a successful, respected figure. This raw, frustrated document, a glimpse of a different reality, felt like a slap in the face. She wanted to crumple it up and throw it away, to pretend she hadn't seen it, but the words seemed to hold her captive.

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