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 email arrived, plain text, from the anonymous email address she’d been corresponding with for months: “I'm at the library, by the fiction section. Come find me.” A grin split across her face, so wide it almost hurt. She scrambled out of bed, ignoring the late hour. A sense of urgency, an electric charge, filled her.

She pulled on the first clothes she found, throwing her hair into a messy bun. The library, usually a sanctuary of quiet, now beckoned with promises. Each step felt lighter. The world, normally muted, seemed filled with color and possibilities.

Emotion: hope

Cluster: Positive / Joy
PC1 (Valence): 3.59 Positive
PC2 (Disposition): -0.12

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de1.065
la0.865
B0.465
🤩0.412
H0.407
Suppressed:
😞-0.459
C-0.451
不-0.449
😣-0.443
S-0.442