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 old woman, her face etched with the stories of a life well-lived, stood clutching a worn photograph album. Her daughter, Sarah, watched her, concern etched on her face. "Mom, are you sure you want to stay?"

The woman simply nodded, her gaze fixed on the familiar silhouette of the house. She remembered the birth of her children within those walls, their first steps, their first words. The sleepless nights, the joyful celebrations, the quiet moments of companionship.

She opened the album, her fingers gently caressing the faded photographs. She saw her late husband’s smiling face, her children’s innocent eyes. Memories swirled around her, a comforting warmth against the coldness of the impending demolition. She smiled a bittersweet smile, a final farewell.

Emotion: sentimental

Cluster: Nostalgia / Reflection
PC1 (Valence): 0.89 Positive
PC2 (Disposition): -1.58

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
L0.677
fondly0.567
memories0.503
曾0.488
reminis0.466
Suppressed:
either-0.515
or-0.488
[]-0.473
--0.408
не-0.396