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).

He stared at his reflection, the lines around his eyes etched deeper than usual. The news had spread like wildfire. A private commission, a sprawling landscape across a woman’s entire back. He felt exposed, stripped bare. The woman was at his art exhibit the day prior, laughing and smiling. Her gesture was now a joke to him.

He picked up a palette knife, running its cold edge across his cheek, a perverse desire to replicate the sensation of pain. He threw it down, the clatter a sharp punctuation in the quiet room. He had poured his heart and soul into his work, and now it was out there, on display, a public spectacle. He wanted to rip all of his work from his gallery, and lock himself away.

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