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 vibrant energy of the wedding seemed to evaporate as soon as she caught sight of him. A chill snaked up her spine, even though the room was warm. Emily's hand went to her throat, as if to physically restrain a sudden urge to flee.

Everything felt… skewed. The decorations, the music, the laughter of the guests - it all felt like an elaborate performance, designed to distract from the central, undeniable reality: He was here.

She gripped her clutch so tightly that her knuckles turned white. A strange hollowness had replaced her previous lightheartedness. She attempted a smile, but it felt brittle and artificial. The happiness of the occasion felt… alien.

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