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 school play program lay open in front of me, listing the names of the cast and crew. My daughter, Maya, was playing a tree. I scanned the page, my eyes lingering on the director's name: Mr. Sterling. The same Mr. Sterling who'd guided me through my own disastrous attempts at performing Shakespeare. A sensation of almost giddy disbelief swirled inside me.

I decided to volunteer for set design, just for an excuse to spend time in the rehearsal space. The stage, the scent of paint, the stage lights – it all transported me. Watching Mr. Sterling direct the children, I felt a deep, almost indescribable emotion.

The way he spoke, the way he gestured – it was all so familiar, so comforting. It was a tangible link to my own past, to the awkward teenager who had dreamed of being an actress. I was overwhelmed. It felt like walking through a dream.

Emotion: awestruck

Cluster: Surprise / Confusion
PC1 (Valence): -0.17 Negative
PC2 (Disposition): -1.20

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
L0.271
suddenly0.260
stranger0.252
plötzlich0.246
wonder0.242
Suppressed:
la-0.450
싶-0.350
K-0.319
rest-0.248
싶은-0.238