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 family had just moved in, and the kids were ecstatic about the huge backyard. One afternoon, while digging a hole for a swing set, their son unearthed something.

It was a small, tarnished metal box. Inside the box were several items. A tarnished silver locket. A photograph of a smiling woman. And a newspaper clipping. The clipping detailed a cold case, a missing person, and an unsolved murder. The address on the clipping was the same as theirs.

The father’s face drained of color as he read the article. He felt a sudden, profound sense of unease. He looked at his children, their faces lit with excitement. He shut the lid of the box quickly. He grabbed his phone. He walked inside.

Emotion: puzzled

Cluster: Surprise / Confusion
PC1 (Valence): -1.17 Negative
PC2 (Disposition): -0.25

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
que0.453
🤔0.421
🤔0.402
🧐0.328
a0.294
Suppressed:
ecstatic-0.265
gladly-0.238
swiftly-0.237
alleviate-0.234
wholeheartedly-0.232