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 antique shop was filled with dust motes dancing in the afternoon sunbeams, a familiar sight for Arthur. He'd been coming here for years, always hoping to find a hidden gem to resell. Mrs. Gable, his neighbor, a woman who always wore the same cardigan and spoke in a breathy whisper, was usually in line with him, haggling over chipped teacups. Today, however, she was purchasing a first edition of "Pride and Prejudice," casually tossing down a wad of cash that would have covered Arthur's mortgage payment for a year. He stumbled back, bumping into a stack of porcelain dolls. His brow furrowed.

He watched her leave, her frail frame somehow radiating an air of untouchable elegance. He felt a tremor start in his stomach. What did she *do*? He’d always assumed she subsisted on a small pension. He found himself staring at his own worn shoes, suddenly feeling terribly inadequate.

Emotion: bewildered

Cluster: Surprise / Confusion
PC1 (Valence): -1.87 Negative
PC2 (Disposition): -0.48

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
будто0.300
ness0.255
像是0.241
worse0.231
Worse0.219
Suppressed:
la-0.373
மகி-0.229
happy-0.226
ecstatic-0.226
K-0.219