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 aroma of freshly brewed coffee, the early morning light streaming through the kitchen window, the comforting weight of his dog, Max, leaning against his leg – it was a moment of perfect tranquility. John sipped his coffee, savoring the bitter, stimulating taste.

His gaze drifted to Mrs. Gable's house across the street. A fleet of moving trucks was parked outside her house. A quick glance inside revealed an elegant interior and several people in business attire. He was surprised; he'd always thought she lived a modest life.

She walked out, in a designer suit he vaguely recognised from a magazine. She smiled at the movers and gave them instructions. A strange feeling, a mix of curiosity and something else, tightened his chest.

Emotion: safe

Cluster: Calm / Serenity
PC1 (Valence): 1.86 Positive
PC2 (Disposition): -2.55

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
soon0.402
affectionately0.357
disfrut0.329
необходи0.325
🥰0.317
Suppressed:
S-0.474
C-0.416
L-0.408
worse-0.379
even-0.370