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 basement was dark and damp, the air thick with the smell of mildew. Chloe stared at the box labeled "Chloe - High School." She’d been tasked with sorting through her deceased mother's things, not this.

She lifted the lid, her hand trembling slightly. She’d come expecting old photos, maybe a forgotten diary. Instead, she found a treasure trove of her adolescent life, preserved in paper and ink. A wave of nausea washed over her. She fought to keep herself composed.

She ran her fingers over a particularly scathing critique of a short story she'd written. She remembered the hurt, the feeling of failure. Now, it was… unsettling. She felt a cold dread creep into her bones. She couldn't understand. Why would her mother do this?

Emotion: shocked

Cluster: Surprise / Confusion
PC1 (Valence): -2.19 Negative
PC2 (Disposition): -0.65

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
будто0.298
😰0.268
sickening0.266
не0.266
que0.257
Suppressed:
enthusi-0.283
开心-0.269
ولة-0.262
ارك-0.257
উৎসাহিত-0.255