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 pawn shop was a cacophony of sounds: the clatter of tools, the murmur of haggling customers. Michael, however, was deaf to them all. His gaze was fixated on one particular case, and a ring nestled inside it. It was like a magnet had pulled him towards it.

He recognized the intricate design immediately. It was his grandfather's ring, a heavy, signet ring that bore the family crest. The ring had vanished after his grandfather's funeral, a fact that had caused considerable family consternation. His head spun.

He found himself walking towards the counter, each step feeling heavy, leaden. He couldn't articulate the feeling that had engulfed him. His hand trembled as he pointed to the ring, the words catching in his throat.

Emotion: mystified

Cluster: Surprise / Confusion
PC1 (Valence): -1.24 Negative
PC2 (Disposition): -0.16

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
que0.458
🤔0.364
🤔0.357
будто0.308
puzzling0.299
Suppressed:
alleviate-0.241
ecstatic-0.232
timeLeft-0.232
rest-0.227
allevi-0.225