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 slammed the file cabinet shut, the metal echoing loudly in the otherwise silent office. His hands were shaking slightly. Mark had a reputation for being unflappable, the cool head in any crisis. But this… this was different. He reread the letter, the crisp paper feeling foreign and unsettling between his fingers. A sister. Someone he’d never known. The news had dropped on him like a lead weight. He rubbed his temples, a dull ache throbbing behind his eyes. He had a presentation in an hour, a crucial one, and all he could think about was this woman, this stranger, suddenly inserted into his life.

Emotion: stressed

Cluster: Fear / Anxiety
PC1 (Valence): -1.37 Negative
PC2 (Disposition): -0.53

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😩0.266
😫0.256
😰0.239
πŸ˜–0.237
😞0.236
Suppressed:
B-0.200
happy-0.197
enjoy-0.176
enjoying-0.172
bahagia-0.169