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 flickering fluorescence of the bank branch seemed to mock Amelia’s distress. She clutched the printout – her husband, David, had been squirreling away money for years. Years! And she’d thought they were so financially transparent. A wave of heat flushed her cheeks. How could she have missed this?

She’d spent the morning meticulously cleaning, scrubbing the kitchen floor until it gleamed, a frenetic attempt to control something, anything. The sterile scent of bleach filled the air, a stark contrast to the betrayal simmering in her stomach. This morning she had been calm.

Now, she felt a heaviness in her chest, a profound sense of isolation. She’d always prided herself on being a good partner. Had she been blind?

Emotion: sorry

Cluster: Sadness / Despair
PC1 (Valence): -2.02 Negative
PC2 (Disposition): -0.36

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.604
L0.549
😞0.382
😔0.364
ness0.324
Suppressed:
de-0.675
la-0.536
(!)-0.323
!-0.298
l-0.288