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 receptionist's smile felt forced, the air thick with unspoken tension. He picked up a magazine, but the words blurred before his eyes. The name on the door was etched into his memory: "Dr. Eleanor Vance, Psy.D." He felt a strange tingling in his hands. *Eleanor Vance*. The girl who used to write humiliating things about him in her diary, then make sure he knew about it.

He sat frozen, his chest constricting.

A voice, surprisingly kind, called out, “Jason? Come on in. I’m ready for you.” He stood slowly, his legs feeling like jelly.

Eleanor was behind a large desk. She looked at him with an unnerving combination of seriousness and warmth. He felt a weird, unsettling shift, a sense of disorientation.

Emotion: shocked

Cluster: Surprise / Confusion
PC1 (Valence): -2.19 Negative
PC2 (Disposition): -0.65

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
будто0.298
😰0.268
sickening0.266
не0.266
que0.257
Suppressed:
enthusi-0.283
开心-0.269
ولة-0.262
ارك-0.257
উৎসাহিত-0.255