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).

“Condo conversion,” the email read, in a chirpy, over-enthusiastic font. She almost deleted it immediately. Instead, she stared at it, her fingers drumming a furious tattoo on her desk. The landlord’s glee practically jumped off the screen.

She’d spent years cultivating her apartment, filling it with things she cherished. It was an imperfect sanctuary, a haven from the relentless chaos of the city. Now, they wanted to rip her from it.

She reread the email, her stomach churning. The offer was insulting. A pittance. A cruel joke. She pictured the smug real estate agent, the shiny brochures, the open houses. She closed the laptop with a snap. She’d show them.

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