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 morning sun glinted off the polished mahogany of his desk. He leaned back in his leather chair, the satisfying creak a soundtrack to his triumph. A steaming mug of coffee sat untouched, the aroma a mere supporting character to the main event. Heโ€™d just finished reading the online edition of *Literary Luminaries Quarterly*, and there it was. His words. His brilliance. Underโ€ฆ *her* name. He chuckled, a low rumble that vibrated in his chest.

He swiveled the chair, the window offering him a view of the sprawling city. He could almost feel the weight of expectation shifting, the pressure of those who had doubted him starting to crumble. He tapped a manicured fingernail against the screen, re-reading the byline. Well, theyโ€™d all find out eventually, wouldn't they? He had a plan. A very good plan.

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
๐Ÿ˜Ž0.570
๐Ÿ˜‰0.564
Suppressed:
S-0.579
own-0.486
๐Ÿ’”-0.483
๐Ÿ˜ž-0.464
๐Ÿ˜ฃ-0.455