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 bell above the door jingled, announcing Liam's arrival. He barely registered it. His gaze was fixed on the ring, gleaming in the display case like a mocking beacon. His grandfather, always a man of principle, would be horrified. This was another casualty of his father's gambling.

He ignored the shopkeeper's greeting, his jaw clenched so tightly it ached. This ring, a symbol of a lifetime of love and commitment, was now a commodity, subject to the whims of the market. The injustice of it all burned within him, a slow, simmering fire.

"How much?" he managed to croak out, his voice a strained whisper. The shopkeeper, a man whose eyes held a flicker of amusement, named a price that seemed to mock the ring’s true value. Liam felt a coldness spread through him. He reached for his wallet.

Emotion: resentful

Cluster: Shame / Guilt
PC1 (Valence): -1.47 Negative
PC2 (Disposition): 1.50

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.485
aggravated0.263
愤0.261
👎0.245
L0.241
Suppressed:
own-0.318
soon-0.225
随后-0.225
serendip-0.219
🤩-0.210