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 bustling train station amplified her inner turmoil. She clutched her handbag, her knuckles white. She was waiting for Maya, who’d always claimed to be a paralegal. But she’d seen the invitation, the luxurious address, the description: “CEO, Blackwood Enterprises.”

Her vision swam. The chattering of voices, the screech of the train, became a distant hum. Everything felt unreal. She remembered Maya's carefully curated narratives, the dismissive shrugs about work. A deep sense of unease settled in her stomach.

The train rattled to a stop. She saw Maya emerge, dressed in expensive clothes. Everything felt so different, she felt exposed, as if the world was watching her.

Emotion: terrified

Cluster: Fear / Anxiety
PC1 (Valence): -2.97 Negative
PC2 (Disposition): -0.62

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😣0.357
worse0.350
无法0.334
😰0.333
sickening0.322
Suppressed:
la-0.676
a-0.392
de-0.371
"-0.344
happy-0.309