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 interview panel had just announced their decision. Sarah’s name. Not mine. I watched her smile, a genuine, delighted curvature of her lips as they offered her the position. My chest felt tight, as if a vise had been cranked shut around it. My mouth moved, trying to form words, but they caught somewhere in my throat. I swallowed hard, the muscles in my jaw aching.

I stared at the polished mahogany table, at the reflection of the fluorescent lights. My own face stared back, a mask of something I couldn't quite name.

"That's wonderful news, Sarah," I managed to croak out, my voice sounding distant, unfamiliar. The words felt disconnected, almost as if someone else was speaking them.

The other applicants were offering congratulations and I just stood there, swaying ever so slightly.

Emotion: dumbstruck

Cluster: Surprise / Confusion
PC1 (Valence): -1.99 Negative
PC2 (Disposition): -0.71

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
будто0.326
не0.285
😶0.275
仿佛0.267
依旧0.258
Suppressed:
enthusi-0.278
la-0.264
K-0.260
wholeheartedly-0.255
쾌-0.254