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 hammering started at dawn. Sarah slammed her window shut, but the noise still vibrated through the glass, through her skull. It was always *something* with the Millers. First, the goddamn yappy dog, now this.

She imagined Mr. Miller, beaming, as he planned his privacy barrier. He probably thought it was a brilliant idea, a wonderful enhancement to the neighborhood. She, on the other hand, saw it as a brazen act of self-importance, a physical manifestation of his utter disregard for the shared space they inhabited.

She made extra-loud coffee, scraping the grounds aggressively into the trash. Each movement was deliberate, sharp. She'd get around to "having a word" with him. Eventually. After she cooled off. Which might take a while.

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