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 musty scent of cedar and regret filled Amelia's nostrils as she pulled the box from under her mother’s bed. Inside, nestled amongst moth-eaten shawls, was a worn leather-bound journal. It wasn't the diary she'd expected, filled with secrets and grievances. This was a list. Bucket List. At the top, scrawled in her mother’s elegant script, was “Learn to Ballroom Dance.” Amelia clenched her jaw. She’d always refused, citing her “two left feet.”

Amelia ran a hand through her hair, a nervous gesture she’d inherited. A photograph fell out. Her mother, laughing, arms around a handsome stranger. They were on a dance floor, mid-tango. Amelia quickly put the photo back. Her stomach twisted, a familiar knot. She’d told everyone her mother hated dancing.

Emotion: ashamed

Cluster: Shame / Guilt
PC1 (Valence): -1.97 Negative
PC2 (Disposition): 1.06

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.376
ashamed0.344
ness0.327
逃0.297
😞0.295
Suppressed:
de-0.539
que-0.300
😍-0.288
(!)-0.269
🤩-0.242