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 weight of the document felt heavy in my hands. My sister, Beatrice, inherited the family's renowned publishing house. I, Finn, was left with a single, handwritten manuscript. The lawyer, his voice smooth, offered the usual platitudes about "legacy" and "family history."

Beatrice was already planning on the future, her eyes shining with ambition. I felt something akin to vertigo, and focused on keeping my posture. I knew something was not quite right.

I clutched the manuscript, its pages filled with a story I didn't recognize. The lawyer’s gaze lingered on me, and then he left. I felt a prickling sensation at the back of my neck. I began to read. The more I read, the more certain I was that I needed to know more.

Emotion: suspicious

Cluster: Suspicion / Vigilance
PC1 (Valence): -0.82 Negative
PC2 (Disposition): 0.82

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
🤔0.420
🤔0.415
🧐0.362
疑0.352
suspicious0.352
Suppressed:
and-0.307
alleviate-0.301
ได้รับการ-0.298
allevi-0.285
mouthful-0.284