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 morning sun warmed Elias's face as he stretched languidly in the kitchen. He hummed a tuneless melody while he waited for the coffee to finish brewing, the aroma of the beans a comforting embrace. His hands moved slowly as he buttered a slice of toast, the simple act a meditation. He’d planned to drive to the farmers market, but that could wait. A slow amble through the garden sounded even better. The birdsong outside was a constant, gentle backdrop to his morning.

He stepped outside and then stopped. Where the old Ford pickup had been, there was nothing. A hollow space stared back at him. A small, white sign fluttered on the grass, a harsh visual disruption. His brow furrowed only slightly. He walked toward it, his gait still slow and unhurried. “Well now,” he said to the empty driveway, a gentle chuckle escaping his lips.

Emotion: peaceful

Cluster: Calm / Serenity
PC1 (Valence): 2.40 Positive
PC2 (Disposition): -2.79

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
прият0.408
disfrut0.407
gently0.390
affectionately0.379
thoughtful0.376
Suppressed:
de-0.541
S-0.493
L-0.396
'-0.388
worse-0.367