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 air in the filing room was thick with dust and the scent of aged paper. Michael was now sorting through old documents. He stared blankly at the stacks, a feeling of listlessness washing over him. The repetitive work felt like a slow, deliberate form of torture.

He picked up a file, his fingers tracing the faded lettering on the cover. He felt a strange sense of detachment, as if he were observing himself from a distance. He found himself absentmindedly rearranging the files, even though there was no need.

He leaned against a filing cabinet, the metal cool against his back. He sighed, the sound echoing in the empty room. He longed for something to shake him out of his torpor.

Emotion: bored

Cluster: Fatigue / Lethargy
PC1 (Valence): 0.04 Positive
PC2 (Disposition): -1.11

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.387
’0.350
unremarkable0.325
दिलचस्प0.323
seemingly0.319
Suppressed:
B-0.592
P-0.304
because-0.292
因为它-0.287
ancar-0.278