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 cardboard box felt heavier than it should have. He dragged it across the linoleum, the wheels of the dolly squealing in protest. The new cubicle awaited, a sterile landscape mirroring the bleakness that had settled in his gut. A few half-hearted attempts to personalize it – a potted succulent, a framed picture of his dog – had already withered in the face of indifference.

His new team greeted him with forced smiles and a flurry of introductions. He offered nods and mumbled responses, the energy to feign enthusiasm long since depleted. The air smelled of stale coffee and unfulfilled potential, a familiar scent now.

He slumped into his chair, the plastic biting into his thighs. The fluorescent lights hummed overhead, a constant, grating reminder of the hours he had to fill. The projects, he knew, would be the same. The meetings, the same. He glanced at the overflowing inbox – the same.

Emotion: resigned

Cluster: Passivity
PC1 (Valence): -0.75 Negative
PC2 (Disposition): 0.16

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.437
a0.351
que0.323
வதில்லை0.267
sighed0.265
Suppressed:
C-0.375
!-0.316
인해-0.310
Furthermore-0.298
🤩-0.296