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 office air conditioning was broken, and the fluorescent lights hummed a maddening tune. Mark rubbed his temples, trying to focus on the spreadsheet in front of him, but his mind kept drifting. He’d just learned he was being passed over for a promotion, a crushing blow. Lily, his sister's daughter, was visiting, and her boundless energy, coupled with the relentless sunshine, was exhausting. He found a drawing on the fridge - a red blob with googly eyes and long, spindly legs. "He protects me from the monsters," Lily had said, her voice full of innocent conviction. The image triggered a memory, the recollection of his own imaginary friend, a red blob with googly eyes and spindly legs. Mark felt his throat close up, a feeling that made it hard to swallow.

Emotion: overwhelmed

Cluster: Fear / Anxiety
PC1 (Valence): -2.52 Negative
PC2 (Disposition): -0.94

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.385
😫0.332
πŸ˜–0.319
😣0.319
😩0.314
Suppressed:
la-0.576
de-0.496
a-0.297
happy-0.290
l-0.264