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 rhythmic clack of the keyboard had been the soundtrack to Leo's existence for the last year. He had barely left his apartment, consumed by his passion project: a revolutionary new operating system. His shoulders were locked, his neck a rigid cable. He was nearly there. He just needed to upload the final files.

He’d almost missed the email. A simple, declarative sentence in the subject line: “Regarding the ‘Phoenix’ project.” His heart began to hammer. The sender was a prominent tech entrepreneur. The email went on to describe the release of a new OS, identical in every detail to his own.

He scrolled through the attached link, his vision blurring. A wave of nausea hit him. He reached for the phone, knuckles cracking. He was unable to move it.

Emotion: tense

Cluster: Fear / Anxiety
PC1 (Valence): -1.32 Negative
PC2 (Disposition): -0.53

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.213
😰0.191
느껴0.183
😣0.183
😖0.181
Suppressed:
a-0.260
de-0.258
la-0.256
happy-0.218
B-0.198