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).

He kept finding the drawings scattered around the house. Mark, a practical man, a man of logic, found himself checking the locks on the doors. Each time, the same creature stared back at him from the paper: a creature with oversized eyes and a perpetually tilted head, always positioned just at the edge of the frame. It was the same creature he'd conjured to combat the monsters under his bed when he was a boy.

“Mia,” he said to his sister, his voice unusually strained. “What’s this all about?”

Mia looked up from her book, unconcerned. “Oh, that’s Fuzzball. Chloe’s new friend.”

A bead of sweat trickled down Mark's temple. He felt a disquiet he hadn't experienced since his university physics exam. He had the unsettling sensation of being watched, of something being slightly off-kilter, like a picture hanging crooked on the wall.

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