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 photographer gestured for me to join the group for a photo. I hesitated, my stomach twisting. I caught a glimpse of him standing near the back, a small smile playing on his lips. I forced myself to walk towards the group, my steps slow and deliberate.

I felt suddenly clumsy, aware of every movement, every misplaced word. The way my hand rested on my friend’s arm felt awkward and rigid. I could almost feel his eyes on me, judging me. My cheeks felt hot. When the photographer said, “Say cheese!” I could barely manage a weak, strained smile.

After the photo, I lingered at the edge of the group, pretending to admire the bride's bouquet. My gaze flicked towards him. He was looking at me. I quickly looked away, my face burning, feeling exposed and vulnerable, like I was under some unseen microscope. I moved away, needing distance.

Emotion: self-conscious

Cluster: Embarrassment
PC1 (Valence): -1.03 Negative
PC2 (Disposition): 0.52

Role in Research

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

Logit Lens (Layer 40)

Tokens promoted/suppressed when the self-conscious vector is projected through the unembedding matrix.

Promoted:
S0.307
C0.298
ness0.273
尴尬0.263
😶0.242
Suppressed:
que-0.399
de-0.266
triumphant-0.266
আনন্দের-0.247
ايا-0.238