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).

My jaw nearly unhinged. Mark, the bane of my existence throughout elementary school, a therapist? The woman at reception had just confirmed it, and my stomach did a series of giddy flips. I gripped the worn leather of the waiting room chair, a wide, goofy grin plastered across my face. I imagined the scenarios: him sweating under pressure, me calmly recounting his playground antics. The possibilities… They were endless.

My fingers drummed a rapid staccato against my thigh, the sound echoing in the otherwise quiet room. I picked at a loose thread on the armrest, pulling it until it snapped. My chest felt tight, in a good way. Like I had swallowed a balloon and it was slowly inflating.

“Mr. Davis?” a voice called. My heart leaped into my throat. This was going to be better than I had ever imagined.

Emotion: thrilled

Cluster: Positive / Joy
PC1 (Valence): 3.31 Positive
PC2 (Disposition): 0.07

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.613
!0.544
B0.508
🤩0.499
🎉0.457
Suppressed:
S-0.601
😞-0.536
ness-0.462
😥-0.351
😔-0.341