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 library was usually her sanctuary, but today it felt like a prison. Sarah hunched over a table, surrounded by the weight of unread books. The accusations of plagiarism hung over her head like a dark cloud.

She traced her finger along the spine of a book, her mind replaying the events that led to her downfall. She should have known better, she should have been more diligent. She felt like such a fool. The words on the page blurred, the sentences twisting into accusations.

Her breath hitched as she fought off a wave of nausea. She grabbed her bag and fled the library, the quiet sanctuary now a place she could no longer bear to be. Sunlight felt piercing on her face.

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