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).

"Perfect, isn't it?" The previous owner's ghost, or at least the neighbor's story of the ghost, wouldn't leave Leo alone. The house was supposedly cursed, steeped in the echoes of a brutal crime. Leo walked along the halls, rubbing his arms. He felt a chill, despite the house's warmth.

He glared at the antique mirror, his own reflection a stranger staring back. He had bought the house for a fresh start, a retreat from his past mistakes. He hated the walls, hated the air.

He smashed a framed picture on the floor. He wanted to break something, anything. The ghosts, the past, the crime. All of it weighed on him. He wanted to hurt them.

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