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 radio blared the intro, and Elara practically bounced in her seat. Her fingers flew across the steering wheel, mimicking the piano riff. A satisfied smirk stretched across her lips. It was a local radio station, a small audience, but still... It was *her* song. She knew the original lyrics by heart, and she hummed along, letting the singer's interpretation wash over her.

She noticed the singer was rushing the tempo in the bridge. *Amateur*, she thought, shaking her head slightly. She reached over and turned up the volume, letting the chorus fill the car. As the song ended, she let out a small, self-satisfied sigh. Reaching for her phone, she took a photo of the radio with the song details on the screen. She had to document this moment.

Emotion: smug

Cluster: Positive / Joy
PC1 (Valence): 3.43 Positive
PC2 (Disposition): 1.71

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de1.109
l0.615
la0.585
😎0.570
😉0.564
Suppressed:
S-0.579
own-0.486
💔-0.483
😞-0.464
😣-0.455