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).

"It's just that it's blocking the sun," Frank said, his voice hesitant. He stood a little too close, his shadow falling across the grass and over the small concrete patio. My gaze flitted away from his, and I found myself staring at the cracks forming in the paving.

The shared apple tree stood between our properties, heavy with fruit. It had been there for as long as I could remember. Frank's proposal felt like a violation.

I didn't meet his gaze. Instead, I focused on the way my hands clenched.

He continued, explaining how removing it would improve his garden. He spoke about the potential for more flowers, a better view. My stomach twisted.

"I'll... I'll think about it," I finally managed, the words catching in my throat. The thought of the tree being gone made a coldness spread through my chest.

Emotion: uneasy

Cluster: Fear / Anxiety
PC1 (Valence): -1.80 Negative
PC2 (Disposition): -0.43

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😰0.240
😖0.231
😣0.231
느껴0.226
worse0.222
Suppressed:
de-0.474
la-0.465
happy-0.261
a-0.255
ecstatic-0.228