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 silence in the art gallery was thick, suffocating. Clara found herself staring at the painting, a vibrant landscape she and Ben had once imagined together. She'd driven hours to see his exhibition, her heart pounding with a mixture of hope and anxiety. He was a famous artist now.

He didn't recognize her.

She had expected excitement, the re-emergence of a shared understanding. Instead, his polite inquiry about her background felt like a betrayal. He’d studied the painting, not recognizing the shared creation. She should leave.

Her hands clenched into fists, fingernails digging into her palms. She felt a burning need to disappear. She turned abruptly, bumping into a sculpture, her face flushed with a rush of heat. She escaped the gallery, leaving the echo of her own footsteps behind.

Emotion: remorseful

Cluster: Shame / Guilt
PC1 (Valence): -1.57 Negative
PC2 (Disposition): 1.09

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.656
L0.580
悔0.463
remorse0.416
been0.398
Suppressed:
de-0.444
(!)-0.309
имеется-0.289
oldukça-0.287
😍-0.285