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 time capsule was hidden in the hollow of an ancient oak, a place Liam often frequented. The discovery felt like an intrusion. He traced the circumference of the metal canister with his fingers, a tightening in his jaw. Why did he not recall this?

His breath hitched. He wasn’t a sentimental person. He didn’t save things, didn’t hold onto memories. Yet, there it was, bearing witness to something he should remember. He cautiously unlatched the lid, the sound echoing in the tranquil woods.

Inside, a single, tarnished silver locket rested. He picked it up, his thumb tracing the intricate carvings. A sense of unease settled into the pit of his stomach. A sense that this locket, and everything within this box, held a secret. He didn’t know what the truth was, but a feeling of profound discomfort began to fester.

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