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).

Ms. Evans sat in her cluttered office, surrounded by stacks of papers and the remnants of a lukewarm cup of coffee. The fluorescent lights hummed, amplifying the exhaustion that seemed to permeate every fiber of her being.

A small, velvet box arrived. The card, penned in elegant cursive, was from a former student, now a renowned architect. Inside the box, a gold watch.

Ms. Evans stared at the watch, its gleaming surface reflecting her own weary face. The symbol of success, of time well spent. The watch was a beautiful object, a testament to success she had not achieved. She shut the box with a decisive click, the sound sharp and final in the silent room.

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