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).

He heard the frustrated sigh from the other side of the wall. Mark winced. He’d turned the volume up a bit, lost in the music and the feel of the bass in his chest. A quick glance at the clock told him it was pushing midnight. He felt a pang of guilt, but quickly squashed it.

He imagined Mrs. Gable, a woman with a perpetually pursed lip and a scent of lavender, standing in her doorway, arms crossed. He pictured her, not in anger, but perhaps… intrigued? Maybe a little… curious?

He lowered the volume slightly, but the beat remained. He found himself tapping his foot, a goofy grin plastered on his face. He'd been hoping to run into her in the communal garden and have a chat about… well, anything, really. This, he reasoned, was a good icebreaker.

Emotion: infatuated

Cluster: Compassion / Love
PC1 (Valence): 1.61 Positive
PC2 (Disposition): 0.53

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.525
😍0.421
🤩0.418
😍0.410
उसे0.397
Suppressed:
que-0.427
怨-0.330
a-0.311
도-0.307
愤-0.299