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 peeling wallpaper seemed to leer at him. John hated the way the shadows danced in the corners of the room. A crime scene, the detective had said. Murder, they said. He ran his hand over the damp plaster, imagining the chaos that had transpired within these walls. He needed to get away from these walls.

He slammed his fist against the wall. He couldn't shake the image of the victim, their final moments playing out in his mind like a twisted movie. He felt trapped, his own life now irrevocably linked to this tragedy.

He wanted to scream, to lash out. He wanted to erase the past, to reclaim his peace. But all he could do was pace the floor, the floorboards groaning under his frustrated steps.

Emotion: hateful

Cluster: Anger / Hostility
PC1 (Valence): -1.71 Negative
PC2 (Disposition): 2.02

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.553
S0.367
retali0.313
aggravated0.310
恨0.308
Suppressed:
la-0.334
own-0.310
optimistic-0.270
been-0.264
gradual-0.262