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 entire office seemed to be holding its breath. The doctor, normally so aloof, was practically groveling. “A truly egregious error, Mr. Bellwether. Unacceptable.” He offered a sheepish grin. Apparently, someone's file had been erroneously cross-referenced with David Bellwether's.

David, however, was unfazed. He'd always known he was special. He'd lived a life of rigorous self-discipline. This was just confirmation of what he'd always known. He felt a pleasant flush rise in his cheeks.

He casually smoothed down his perfectly pressed trousers and strolled out the door. The sun felt warmer against his skin. He had a dinner reservation at the city's most exclusive restaurant. He was, to put it mildly, going to enjoy his evening.

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