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).

She hunched over the dusty blueprints, the paper crinkling under her nervous grip. Amelia traced a shaky finger across the lines, feeling the weight of the past pressing down. Each day bled into the next, a monotonous cycle she navigated with a leaden heart. The new neighbors, a cheerful couple, had brought over a housewarming gift, a basket of organic produce. She'd mumbled a thank you, retreating to the solitude of the study. A cold dread had wrapped around her when the police revealed the previous owners had been embroiled in a bitter feud. The final act was played out within her very walls. Her existence felt thin, stretched and easily broken.

Emotion: worthless

Cluster: Sadness / Despair
PC1 (Valence): -2.10 Negative
PC2 (Disposition): -0.15

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
L0.427
S0.405
😞0.386
ness0.339
缺乏0.326
Suppressed:
de-0.320
(!)-0.286
随后-0.268
😍-0.262
tc-0.255