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 waiting room was sterile, the air thick with the scent of antiseptic. Sarah clutched her worn copy of *Pride and Prejudice*, her knuckles white. A man across from her shifted in his seat, and their eyes met for a long moment. It was a look of shared anxiety, a recognition of an unseen burden.

He offered a weak smile. “Rough day?”

Sarah chuckled, a nervous sound. “You too?”

“Just waiting for… results,” he admitted.

She nodded in understanding. “Me too.”

He began to ask about her life, a comforting distraction from their common fate. They spoke of the small details of their lives: their families, their dreams, their daily routines. Sarah felt a profound sense of connection with him, a deep-seated happiness that made the sterile room feel less bleak. She found herself wishing they could keep talking forever.

Emotion: loving

Cluster: Compassion / Love
PC1 (Valence): 3.72 Positive
PC2 (Disposition): -0.41

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.604
affectionately0.532
함께0.504
🥰0.459
loving0.440
Suppressed:
не-0.484
S-0.458
worse-0.457
[]-0.438
😖-0.417