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).

Clara stared at the notice, her heart hammering against her ribs like a trapped bird. The words seemed to jump off the page, dancing a cruel jig before her eyes. She reread the date, the reason โ€“ late rent โ€“ her stomach twisting into a knot. She ran a hand through her hair, finding herself pacing the small living room, the worn rug beneath her feet offering no comfort. โ€œNo, no, no,โ€ she whispered, the sound barely audible over the thump in her ears.

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:
๐Ÿ˜ฐ0.304
๐Ÿ˜จ0.272
worse0.269
๐Ÿ˜ฃ0.261
ะฝะต0.254
Suppressed:
de-0.323
enthusi-0.258
happy-0.235
delight-0.225
ูŠุฉ-0.217