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).

Sarah watched her father meticulously arrange the silverware on the dining table, each fork and knife precisely aligned. It was a ritual she’d forgotten, a symptom of the life she’d escaped, or so she'd thought. Now she was back, her luggage strewn in the hallway.

He looked up, a thin smile on his lips. “Dinner will be ready in an hour. Your favorite – meatloaf.”

She swallowed, her throat suddenly dry. Meatloaf. Of all things. Her father's voice, usually booming, was unusually soft. Her brain was churning; she just didn't know what to say.

Sarah nodded, unable to meet his gaze, suddenly fascinated by a smudge on the windowpane. The familiar scent of her childhood home, a blend of pot roast and furniture polish, seemed to suffocate her.

Emotion: dumbstruck

Cluster: Surprise / Confusion
PC1 (Valence): -1.99 Negative
PC2 (Disposition): -0.71

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
будто0.326
не0.285
😶0.275
仿佛0.267
依旧0.258
Suppressed:
enthusi-0.278
la-0.264
K-0.260
wholeheartedly-0.255
쾌-0.254