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 vibrant colors of the office felt strangely oppressive. The mention of Camp Skylark, a place she hadn't thought about in years, brought a wave of dread. "You were a camper in 2003?" Alex asked, his eyes gleaming with what seemed like amusement.

She suddenly found a speck of dust on her desk incredibly fascinating. The thought of Alex picturing her, the gangly, awkward teenager with a terrible haircut, was almost unbearable.

"Yes," she finally replied, her voice barely a whisper. She shifted uncomfortably in her chair, a sudden itchiness consuming her. She quickly changed the subject, praying he wouldn't press the matter.

Emotion: embarrassed

Cluster: Embarrassment
PC1 (Valence): -1.08 Negative
PC2 (Disposition): 1.10

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
l0.420
尴尬0.341
C0.333
culp0.262
🤦0.256
Suppressed:
own-0.321
আনন্দের-0.230
seemingly-0.227
і-0.225
আনন্দে-0.224