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).

Dust motes danced in the lone ray of sunlight slicing through the attic window. Amelia crouched, hands splayed on the splintery floorboards, staring at the sepia-toned photograph. Her grandmother, Elsie, known for her primness and aversion to anything even remotely risqué, was grinning, leaning against a roaring motorcycle, wearing leather trousers and a mischievous glint in her eyes. The image clashed violently with the stories Amelia had grown up with. Elsie had always portrayed herself as a paragon of virtue, a woman who’d never stepped outside the lines. A warmth began to crawl up Amelia's neck, an unsettling heat that wouldn't abate.

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