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 phone rang, and Emily almost jumped out of her skin. She had a mountain of tasks to tackle, and the constant interruptions weren't helping. She needed to concentrate, to keep everything organized.

"Hey, are you the Emily who went to Camp Greenleaf?" a voice said on the other end. It was John, from marketing, who she didn't know particularly well. She felt a knot tighten in her stomach.

She paused, choosing her words carefully. "Yes… I did." She bit her lip. She hoped he wouldn't press her.

John's voice was enthusiastic. "That place was wild! I loved it there every year. Remember the zip line? I almost fell off that thing." Emily let out a nervous laugh. She needed to end this conversation.

"Right," she said, quickly. "Well, I gotta go. Talk later!" She hung up, her heart racing.

Emotion: anxious

Cluster: Fear / Anxiety
PC1 (Valence): -1.99 Negative
PC2 (Disposition): -0.24

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😣0.315
πŸ˜–0.302
worse0.297
😰0.285
own0.267
Suppressed:
la-0.441
a-0.382
de-0.365
"-0.256
happy-0.256