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 waiting room smelled of antiseptic and something vaguely floral, like a funeral home. Mark tapped his foot, a restless rhythm against the polished floor. He stared at the name on the door: Dr. Evelyn Reed, LCSW. Evelyn Reed. The girl who used to steal his lunch money and call him "Four Eyes." Now, she had a doctorate. Mark felt a knot tighten in his stomach.

He fidgeted with the worn strap of his briefcase. The career he'd chosen, a decent enough one in accounting, suddenly felt…flat. Insignificant. He ran a hand through his thinning hair. A wave of heat washed over him as he recalled her, the way she'd always seemed so effortlessly confident, so certain of herself.

The receptionist, a woman with a kind smile, beckoned him. "Dr. Reed will see you now." He stood, the metal of his chair feeling cold beneath him. He took a deep breath, trying to steady the frantic beating of his heart.

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