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 wafted from the diffuser in Emily's living room. She was arranging dried flowers in a vase, their muted colors complementing the soft light filtering through the sheer curtains. The room felt curated, peaceful, a space she had made her own.

A woman arrived, a real estate agent with a weary expression. "I have some... unpleasant news," she began, her voice tight. "About the previous owners. Their lives were complicated."

Emily paused, arranging a particularly delicate sprig of lavender. She considered the words, their impact seeming to graze her like a light wind. She asked the woman to sit, a gentle hand movement inviting her to rest. She returned to her task, the fragrant blossoms a silent, comforting presence.

Emotion: calm

Cluster: Calm / Serenity
PC1 (Valence): 2.72 Positive
PC2 (Disposition): -1.82

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
मुस्कुरा0.415
enjoying0.405
leisurely0.388
:)0.382
কিছুক্ষণ0.382
Suppressed:
S-0.527
L-0.499
worse-0.454
even-0.422
😣-0.420