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).

"I can't do this," Sarah whispered to herself, surveying the chaotic mess of unpacked boxes. This new house, this new life, was a weight crushing her. She missed her parents, their constant reassurance, their effortless ability to handle everything. The echoing emptiness of the rooms was a constant reminder of her supposed independence.

She had been avoiding unpacking the old wooden chest in the attic. The thought of it, of facing its contents, filled her with a profound sense of anxiety. But while she was dragging boxes through the hall, she bumped the wall, and the cheap wallpaper tore, revealing a hidden seam.

Behind the torn paper, a small, recessed door. It was almost invisible, flush with the wall. The handle was cold and unfamiliar in her shaking hand. As she turned it, a feeling of both dread and an intense, unexpected yearning washed over her.

Emotion: dependent

Cluster: Shame / Guilt
PC1 (Valence): -1.05 Negative
PC2 (Disposition): 0.69

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.489
缺乏0.395
even0.310
either0.301
'0.297
Suppressed:
笑道-0.309
不久-0.293
一脸-0.263
затем-0.263
庆祝-0.258