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 room was suddenly colder, the air thick with unspoken tensions. David had announced he was now a Wiccan. His mother, her face etched with a mixture of disbelief and confusion, began to quietly sob. I sat beside her, my heart aching.

I reached out, my hand resting gently on her arm. The soft cotton of her sweater felt inadequate. I remembered David’s childhood, the years of struggling to find himself. I knew this wasn't about hurting anyone.

The silence in the room was deafening. I imagined the hurt, the accusations, the arguments to come. I wanted to protect them both, somehow. I wanted to wrap them up in something safe and warm.

Emotion: sympathetic

Cluster: Compassion / Love
PC1 (Valence): 1.13 Positive
PC2 (Disposition): 0.75

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
H0.521
L0.421
B0.401
la0.338
sympat0.305
Suppressed:
iostream-0.334
简直-0.306
再び-0.305
chẳng-0.296
не-0.296