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 neon sign of "Sammy's Dive Bar" always beckoned to him with a familiar light. The dim lighting, the familiar faces, the clinking glasses, the blues music that he loved – it was his haven. After a long day, he could always find solace there.

The bartender, Mark, was wiping down the counter with a grim face. "Sammy's is done," Mark said, his voice flat. "Last call's on Saturday."

A wave of nausea washed over him. He felt as though he was falling, the floor tilting beneath his feet. The familiar music, usually a comfort, now sounded distorted, melancholic. The dimly lit room, which had always felt intimate, suddenly felt vast and empty.

Emotion: safe

Cluster: Calm / Serenity
PC1 (Valence): 1.86 Positive
PC2 (Disposition): -2.55

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
soon0.402
affectionately0.357
disfrut0.329
необходи0.325
🥰0.317
Suppressed:
S-0.474
C-0.416
L-0.408
worse-0.379
even-0.370