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 cool water enveloped her, the rhythmic pull and glide of her strokes a perfect meditation. Sarah felt at home in the pool. The chlorine scent, the echoing splashes, the feeling of effortless movement – it was her sanctuary. She was a butterfly specialist, her form elegant, her performance consistent. She knew she was good.

Her coach, a stern woman with a perpetually wet ponytail, called her over, her face set. "Sarah, we need you to focus on the breaststroke."

Sarah blinked, the water still clinging to her eyelashes. Her heart sank slightly.

"But... butterfly is what I know," she started to say, her voice barely above a whisper.

Emotion: content

Cluster: Calm / Serenity
PC1 (Valence): 2.98 Positive
PC2 (Disposition): -2.19

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.446
affectionately0.435
😊0.392
😊0.386
thoughtful0.386
Suppressed:
S-0.682
L-0.537
C-0.485
😣-0.362
est-0.350