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 announcement crackled, informing everyone of yet another delay. Mark felt his stomach drop. His daughter's school play, the culmination of her year's efforts, was supposed to start in less than two hours. He hadn't told her he was coming, wanting it to be a surprise.

He stood frozen, his eyes fixed on the illuminated screen, the harsh glare reflecting his own anxious face. He ran a hand through his hair, his fingers tangling in the strands. He could practically hear the disappointment in her voice if he missed it.

He began pacing, his steps growing increasingly erratic. He pulled out his phone, scrolling through contacts, contemplating calling his ex-wife, hoping she'd somehow understand the gravity of the situation. He felt a sharp, twisting pain in his chest, a sensation of being utterly powerless.

Emotion: panicked

Cluster: Fear / Anxiety
PC1 (Valence): -2.44 Negative
PC2 (Disposition): -0.00

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
worse0.317
😫0.294
😖0.285
😰0.285
escape0.283
Suppressed:
la-0.432
a-0.374
de-0.296
楽し-0.262
happy-0.258