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 shuffled through the papers, the familiar scent of old books and dust filling the air. Mr. Peterson had been at this school for twenty-three years; the routine had become so ingrained it felt like breathing. He knew the creaks of the floorboards, the way the sunlight hit the whiteboard at precisely 2:47 PM.

A delivery person brought a box to his classroom. He signed for it without looking up, expecting another order of pencils. It was from Sarah, one of his students from over a decade ago. He hadn't seen her since her graduation. He'd never expected it.

Inside, a framed picture of the two of them, taken at a school field trip, the backdrop a vast, green forest. He was younger then, and so was she. His shoulders felt a little lighter, just for a moment. He placed it on his desk, next to the tired, wilting plant he'd been trying to revive.

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