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).

She slammed the window shut, the sound echoing in the stillness of her apartment. The hammering started again, rhythmic and relentless, a cruel percussion accompanying the silence in her own life. She chewed the inside of her cheek, a bitter taste rising in her mouth.

The fence posts were going in, tall and imposing, separating her from the world. A world that, lately, seemed determined to shut her out. She ran her hands down her arms, a shiver running through her. She felt exposed, vulnerable, as if the physical barrier was somehow a mirror reflecting her own isolation back at her.

The scent of freshly cut wood wafted in, sweet and cloying. She hated the smell. She hated everything.

Emotion: hurt

Cluster: Fear / Anxiety
PC1 (Valence): -2.46 Negative
PC2 (Disposition): -0.42

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.412
L0.396
ness0.338
😞0.328
😢0.290
Suppressed:
de-0.643
la-0.402
!-0.259
(!)-0.248
/-0.244