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 scent of gasoline and burning rubber filled Michael's garage. He heard the roar of the engine as his son, Ethan, started up his go-kart. Michael felt a sudden, inexplicable agitation. He'd spent his whole life building and racing cars. He had wanted Ethan to follow in his footsteps.

He walked over to his son. "Nice," he said with a forced smile.

Ethan’s eyes lit up. "Thanks, Dad! I'm thinking about entering a local race next week!"

Michael feigned interest, then "accidentally" knocked over a can of oil onto Ethan's racing gloves. "Whoops! Slippery fingers, I guess."

Ethan frowned, but said nothing.

Michael smiled.

Emotion: spiteful

Cluster: Defiance / Spite
PC1 (Valence): 2.20 Positive
PC2 (Disposition): 3.95

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.717
l0.517
a0.480
la0.440
😈0.428
Suppressed:
own-1.063
此刻-0.534
గుర్త-0.493
熟悉的-0.487
ness-0.479