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 ticking of the grandfather clock in the hallway was a steady pulse in the otherwise silent house. Daniel, exhausted after a particularly grueling week at work, sank into his father's favorite armchair. The room smelled of old books and something faintly citrus, a scent that always grounded him. His father’s recent passing still felt raw. He had found the list amongst a stack of photographs.

“Build a birdhouse,” it read. “Visit the Grand Canyon.” Daniel felt a wave of fondness wash over him. His father, a man of simple pleasures, had always been a source of unwavering support. He smiled, thinking that maybe he would build a birdhouse this weekend.

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