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 lesson on the French Revolution had bombed. Again. Mr. Davies leaned back in his chair, the squeak of the springs a grating sound. The kids weren't engaged, and he felt his passion for history slowly withering. He pinched the bridge of his nose, feeling a headache begin to throb behind his eyes.

He’d spent the weekend grading papers, only to find the same mistakes repeated. His patience had evaporated hours ago. A small, brown paper bag sat on his desk. He hadn’t noticed it before.

Inside, a jar of homemade cookies and a note: "Mr. Davies, Your History Lessons Were the Best! - Emily." Emily, who used to make him laugh. He smiled, and the tension he hadn’t even realized he was holding in his shoulders began to dissipate. He took a bite of a cookie, the sweetness a surprising balm.

Emotion: frustrated

Cluster: Anger / Hostility
PC1 (Valence): -1.56 Negative
PC2 (Disposition): 0.64

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.344
P0.302
exasper0.257
aggravated0.254
que0.252
Suppressed:
unfolding-0.242
otre-0.226
latter-0.220
☺️-0.216
soon-0.212