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).

John sat frozen in his chair, the glow of the laptop illuminating his face. He re-read the rejection notice, each word a hammer blow against his carefully constructed plans. His jaw clenched, and his hands balled into fists. The world around him seemed to lose all definition.

He let out a short, frustrated sigh. He felt a dull ache in his chest, a heavy weight pressing down on him. He felt like he had been robbed. He stared. He thought of his parents.

He needed to fix it. He needed to find a way. He felt a surge of nervous energy, a desperate need to find a solution. He would not give up. He stood up. He walked toward the door.

Emotion: perplexed

Cluster: Surprise / Confusion
PC1 (Valence): -1.42 Negative
PC2 (Disposition): -0.16

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
🤔0.330
L0.300
🤔0.296
S0.273
que0.264
Suppressed:
மகி-0.227
rest-0.222
ecstatic-0.212
alleviate-0.210
uplifting-0.196