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 scent of lavender, his grandmother's favorite, still clung to the towels in the bathroom. He inhaled deeply, the fragrance triggering a wave of images—his grandmother, her hands gnarled with age, carefully folding the linens. He felt a soft, warm feeling bloom in his chest.

He found himself wandering through the house, touching things: the smooth wooden banister, the worn rug in the hallway. He realized that this was the same house, but it was also a different one. The people were the same but their lives had changed. He felt a sudden, sharp ache of missing her.

His mother, her face etched with worry, watched him from the doorway. He managed a weak smile, wishing he could somehow share this moment with his grandmother.

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