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 email notification popped up, and Ethan immediately knew what it was. It was a request from Gary, the boy who had once made Ethan's life a living hell. Now, Gary’s profile was a portrait of success: a thriving business, a loving family, a life of apparent ease and prosperity. Ethan’s jaw clenched involuntarily.

He was sitting in his sparsely furnished office, staring at a spreadsheet filled with depressing numbers. His business was failing. He had invested everything he had, and it wasn't working. He felt a cold dread creep into his bones. His own life felt bleak in comparison.

He glanced at Gary’s profile, a testament to his own perceived failures. The happiness, the wealth, the validation. It was all a painful reminder of what Ethan lacked. He felt a tightness in his chest, a sensation of being unable to breathe. He wanted to scream.

Emotion: envious

Cluster: Shame / Guilt
PC1 (Valence): -0.65 Negative
PC2 (Disposition): 1.58

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.550
H0.431
B0.406
L0.403
envy0.365
Suppressed:
再び-0.262
мед-0.258
再次-0.246
own-0.239
พร้อม-0.231