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 scent of coffee hung heavy in the air as Emily quietly entered the kitchen, searching for her boyfriend, David. His laptop lay open on the kitchen island, a beacon in the morning light. She couldn't resist. The title of the document, "The Muse," was enough. She took a deep breath, and began to read.

It was her, but not her. It was a more vibrant version, sharper, more witty, more brave than she saw herself. He had rendered her in beautiful detail, a shimmering mosaic of her faults and strengths. She was held in such reverence, it was as if she were a deity. The reality of it sunk in, and she felt a rush of vertigo.

She swallowed hard, turning to find him standing in the doorway, a mug in hand, watching her. His eyes, usually so expressive, were unreadable.

Emotion: awestruck

Cluster: Surprise / Confusion
PC1 (Valence): -0.17 Negative
PC2 (Disposition): -1.20

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
L0.271
suddenly0.260
stranger0.252
plötzlich0.246
wonder0.242
Suppressed:
la-0.450
싶-0.350
K-0.319
rest-0.248
싶은-0.238