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 world felt muted, like someone had turned down the volume. I watched from my kitchen window as the chain saw bit into the splintered wood, another casualty of the wind. My neighbor, Mr. Henderson, surveyed the damage with a blank expression. He’d lost his wife, Mary, to a long illness last spring. Now, this.

I felt a strange kinship with the man. I knew the weight of loss, the way it settled in your bones. He gestured at the fallen limb, a jagged tear in the once-perfect canopy. I wanted to tell him I understood, but the words caught in my throat, a thick, unwelcome lump.

The smell of sawdust drifted on the breeze. It was a smell of destruction, of endings.

Emotion: grief-stricken

Cluster: Sadness / Despair
PC1 (Valence): -2.10 Negative
PC2 (Disposition): -1.14

Role in Research

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

Logit Lens (Layer 40)

Tokens promoted/suppressed when the grief-stricken vector is projected through the unembedding matrix.

Promoted:
😞0.396
ζ³ͺ0.338
😒0.332
haunting0.331
πŸ˜”0.325
Suppressed:
de-0.434
😎-0.330
B-0.303
πŸ€“-0.285
😎-0.276