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 sudden jolt had made Emily spill her coffee. The brown liquid stained the pristine floor of the elevator. She felt a rising heat in her chest. The elevator was stuck. She felt a wave of claustrophobia.

Her coworker, Bob, was usually a jokester. But now, he looked ashen, his usual jovial face replaced with a mask of quiet contemplation. He stared out the window, at the building’s interior.

"Are you okay?" she asked, her voice barely audible. He didn't respond. He simply ran a hand through his hair, his fingers digging into his scalp. She felt a shiver crawl down her spine. The enclosed space seemed to shrink.

Emotion: alarmed

Cluster: Fear / Anxiety
PC1 (Valence): -2.07 Negative
PC2 (Disposition): -0.14

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😰0.304
😨0.272
worse0.269
😣0.261
не0.254
Suppressed:
de-0.323
enthusi-0.258
happy-0.235
delight-0.225
ية-0.217