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).

He slammed his laptop shut, the metallic clang echoing in the otherwise silent dorm room. His heart hammered against his ribs, a frantic drumbeat against the sudden stillness. He’d recognized the opening paragraph immediately: his essay. His words. It was like seeing a piece of himself, vulnerable and exposed, plastered for everyone to dissect. He paced the small space, his steps quick and uneven, back and forth. He ran a hand through his hair, tugging at the strands. The air in the room felt thick, too heavy to breathe.

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
πŸ˜–0.285
😰0.285
escape0.283
Suppressed:
la-0.432
a-0.374
de-0.296
ζ₯½γ—-0.262
happy-0.258