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).

He stared at the screen, the cursor blinking mockingly. Sarah Miller, the name emblazoned above the little "Confirm" button, brought back a flood of memories: a red-faced, freckled tormentor, a voice that could curdle milk. His gut clenched, a familiar, unwelcome sensation.

He leaned back in his chair, running a hand through his already tousled hair. The office was quiet, the fluorescent lights humming a dull song. He had a deadline looming, but somehow, this small digital gesture seemed more pressing.

He finally clicked the accept button, then closed his eyes. The tension gradually faded. He took a deep breath. She could wait. He could wait. He’d deal with whatever he wanted later.

Emotion: patient

Cluster: Passivity
PC1 (Valence): 2.71 Positive
PC2 (Disposition): 0.40

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
rest0.467
eventually0.415
leisurely0.369
慢慢0.357
a0.355
Suppressed:
无法-0.432
even-0.428
인해-0.421
無法-0.415
😣-0.391