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).

"So, you're saying I'm not pregnant?" Liam repeated, the words tasting like ash in his mouth. He'd been ecstatic, then terrified, then… this. The doctor, bless her, looked mortified. "It was a mix-up, Mr. Peterson," she stammered, flipping through the files again. "We are so incredibly sorry."

He wanted to scream. He wanted to throw something. He just stared at the ceiling, the sterile white a stark contrast to the colorful dreams he’d been building over the past week. His partner, Sarah, would be heartbroken. He felt a profound, heavy weight settle in his chest.

“Right,” he finally said, his voice flat, devoid of emotion. “Well. That’s just… great.” He rose, a slow, deliberate movement. He didn't thank the doctor, didn't offer a farewell. He just walked out, the air feeling thick and heavy, as if he had to wade through treacle.

Emotion: scornful

Cluster: Anger / Hostility
PC1 (Valence): -0.76 Negative
PC2 (Disposition): 2.07

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
l0.406
C0.391
S0.276
I0.268
愤0.243
Suppressed:
own-0.396
unexpectedly-0.288
Suddenly-0.260
soon-0.256
serendip-0.256