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 house echoed with the emptiness of absence. Sunlight streamed in through the windows, illuminating dust motes dancing in the air, a constant reminder of how long she'd been neglecting everything. She was emptying the linen closet when she uncovered a small, decorated box with a red bow and her mother's handwriting. Inside, nestled on tissue paper, was the bucket list.

“See a Broadway show,” read the first entry. “Learn to paint.” “Go on a hot air balloon ride.” A heavy weight settled in her chest. Her mother had always been the energetic one, the life of every party.

I felt like an automaton. Every movement felt like wading through molasses. The list seemed to mock her, a reminder of all the things she wasn't doing, all the joy she wasn't experiencing.

Emotion: depressed

Cluster: Sadness / Despair
PC1 (Valence): -2.06 Negative
PC2 (Disposition): -1.08

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😞0.417
每天0.353
无法0.330
own0.311
😔0.309
Suppressed:
de-0.492
B-0.393
P-0.352
l-0.310
!-0.307