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 manuscript lay open on her desk, the pages whispering to her in a language she didn’t understand. Her entire existence, distilled and packaged for others to consume. She felt a profound weight settle in her stomach, a leaden feeling she couldn't shake. Her fingers traced the words, a futile attempt to erase the truth they held.

She pushed herself away from the desk. The grand, sunlit room felt suddenly confining. The ornate furniture, usually a source of pleasure, now seemed alien and cold. She walked to the window, staring out at the vibrant cityscape, but all she saw was a blur.

A deep sigh escaped, shaking her shoulders. This intrusion... it felt like a betrayal.

Emotion: sad

Cluster: Sadness / Despair
PC1 (Valence): -2.52 Negative
PC2 (Disposition): -1.50

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😞0.436
😔0.405
S0.384
无法0.341
L0.337
Suppressed:
de-0.749
la-0.614
l-0.471
/-0.350
!-0.322