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).

Rain clouds threatened, mirroring the mood as David surveyed the haphazard display of discarded belongings. He hadn’t expected to see anything remotely interesting, let alone significant. A weathered teddy bear, remarkably similar to the one he had held dear as a child, caught his eye. The bear’s one remaining eye seemed to stare directly at him.

He circled the table, feigning interest in a cracked teacup, but his gaze kept returning to the bear. The seller, a teenage boy with a bored expression, finally noticed him. “Five bucks for the bear,” the boy mumbled. “Cheap.” David approached, his breath catching slightly. He felt a weird, unsettling lightness in his chest. A trick of the light? Or something else? He wasn't sure.

Emotion: skeptical

Cluster: Suspicion / Vigilance
PC1 (Valence): -0.02 Negative
PC2 (Disposition): 1.45

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.468
que0.396
la0.346
a0.333
🧐0.327
Suppressed:
再び-0.304
终于-0.299
再次-0.270
終於-0.263
momentarily-0.255