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 elevator doors chimed open on the fourth floor, and Brenda practically bounced out. Her new office, the corner suite, was practically calling her name. She hadn't even unpacked her first box, but already she knew she'd be running circles around the drones in marketing. She adjusted the silk scarf at her neck, a slight, self-satisfied smile playing on her lips. Those presentation skills she honed in sales would be wasted here; they'd be her secret weapon.

Her predecessor, the hapless Mr. Henderson, was in the process of clearing his desk. She caught his eye, and, with a curt nod and a widening of her own smile, she sailed past. She could almost taste the promotion already.

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