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 flyer, tucked under the door, landed on her welcome mat with a dull thud. Maria picked it up, her brow furrowed in concentration, as if trying to decipher a foreign language. Her apartment, her beloved space with its chipped paint and character, was going…what? She ran her fingers along the smooth paper, a shiver tracing up her spine.

She clutched the flyer like a lifeline and moved through her space, each object suddenly imbued with a new significance. The old, mismatched furniture, the crooked paintings, even the overflowing bookshelf, suddenly took on a new light. Her jaw dropped. “No way,” she mouthed. This felt like some kind of a joke. The air in her apartment felt suddenly thick, heavy. She felt a lightness, as if she could float away.

Emotion: amazed

Cluster: Surprise / Confusion
PC1 (Valence): 0.74 Positive
PC2 (Disposition): -0.51

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
🤩0.301
de0.289
L0.284
😳0.251
அதிசய0.236
Suppressed:
ing-0.332
l-0.253
😞-0.240
unsuccessful-0.237
unsuccessfully-0.233