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 doorbell chimed, and Sarah’s stomach lurched. It was her sister, visiting with her two perfect children. Sarah plastered a smile on her face, but it felt brittle. She watched, a silent observer, as her sister effortlessly navigated the demands of motherhood, the children’s laughter echoing through the house.

Her mother offered her a cup of tea, a gentle gesture that felt like a reminder of her own dependency. Sarah took a long sip, pretending to enjoy the warmth, while her internal critic chipped away at her. *Look at you*, it whispered. *Still here. Still dependent.*

She excused herself to "check emails," fleeing to the relative privacy of her childhood room, which, with the old stuffed animals and posters, felt even more childish than before.

Emotion: self-critical

Cluster: Shame / Guilt
PC1 (Valence): -1.66 Negative
PC2 (Disposition): 1.28

Role in Research

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

Logit Lens (Layer 40)

Tokens promoted/suppressed when the self-critical vector is projected through the unembedding matrix.

Promoted:
L0.748
S0.662
ness0.425
ly0.374
탓0.351
Suppressed:
la-0.492
soon-0.304
(!)-0.301
secured-0.284
cautiously-0.280