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 glowing screen reflected Sarah's face, making her look gaunt and tired. She scrolled through the hidden profile, each post a fresh stab of dismay. Her daughter, Emily, the shy girl who mumbled greetings and kept her head down, was a vibrant, confident personality online. Sarah felt a twisting in her stomach, a heavy weight pressing on her chest.

She closed the laptop with a snap, the sound echoing in the silent house. Emily would be home from school any minute. She pictured Emily's face, the slight flush of her cheeks, the way she tucked a stray strand of hair behind her ear. How could she have missed this?

The silence felt thick, suffocating. Sarah walked to the kitchen, poured herself a glass of water, and stared out the window. The setting sun cast long shadows across the lawn. A chill settled in the air, mirroring the coldness that had crept into her heart.

Emotion: regretful

Cluster: Shame / Guilt
PC1 (Valence): -1.62 Negative
PC2 (Disposition): 0.76

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
L0.801
S0.774
H0.513
悔0.479
been0.473
Suppressed:
de-0.407
--0.320
随后-0.317
(!)-0.310
oldukça-0.301