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).

"Impossible," the gallery manager was sputtering. "The contract... the release forms…" Maya watched them from the back of the room, her arms crossed, a small, subtle twitch at the corner of her mouth. The manager was struggling.

The exhibit, "Fragments," a series of intimate portraits, was packed. Every viewer gave some kind of reaction. Every one of them was looking at her work. Maya took a small step forward.

She let her eyes drift over the crowd. She felt… buoyant, light. The weight that had been on her shoulders for years had vanished, like a bad dream. She touched the frame on one of the portraits. It had been her dream.

Emotion: satisfied

Cluster: Positive / Joy
PC1 (Valence): 3.20 Positive
PC2 (Disposition): -1.31

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.518
happy0.465
enjoy0.439
relaxed0.430
enjoys0.429
Suppressed:
S-0.515
C-0.511
😣-0.510
worse-0.477
😖-0.453