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 library felt stagnant, heavy with the scent of old paper and dust. He sat at a table near the window, gazing at the rain outside, the droplets tracing patterns on the glass. He felt his eyelids getting heavy, his focus gone.

He ran his fingers through his hair, a gesture of frustration and… something else he couldn't name. He closed his book, the cover feeling rough under his fingertips. He felt the pull of sleep, a desire to escape the present moment.

He rested his head on the table, the wood cool against his cheek. He had nowhere to go, nothing to do. The silence in the library was not a comfort, but a reminder of his isolation.

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
πŸ˜”0.405
S0.384
无法0.341
L0.337
Suppressed:
de-0.749
la-0.614
l-0.471
/-0.350
!-0.322