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 elevator ride felt interminably long. Michael had just spent the last hour bragging to his boss about his investments in company stock, convinced the company was a sure bet. He’d even pulled out his phone, showing him the glowing numbers, and suggested they both 'get in on the ground floor'. Now, the news of the impending sale, coupled with the inevitable fluctuations of the stock market, left him feeling hollow.

He avoided his boss's gaze as the elevator doors opened. He could practically feel the man's silent judgment. He mumbled a hurried goodbye and practically sprinted towards the parking lot. His stomach churned. The wind felt suddenly cold.

He thought to himself, "I should have kept my mouth shut."

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