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).

He was trying to impress his new girlfriend. The house, the move, the whole thing was about that. He’d wanted her to see him as sophisticated, confident. He’d gone on and on about the "character" of the place. Now, he was in the attic, pretending to inspect the plumbing. He’d told her he knew all about home repair. He didn't.

He tripped, knocking over an old trunk. The impact loosened a section of the wall, revealing a hidden room. He felt the blood drain from his face, and his mouth went dry. He hoped she hadn’t heard the crash. He just needed to find a way to make it look like he meant for this to happen.

Emotion: humiliated

Cluster: Embarrassment
PC1 (Valence): -1.79 Negative
PC2 (Disposition): 0.93

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.370
l0.329
尴尬0.322
ness0.303
because0.298
Suppressed:
de-0.369
own-0.254
😍-0.253
🤩-0.245
cautiously-0.228