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 musty smell of the storage unit enveloped Leo. He’d promised his mother he’d go through his grandmother's things. He reached a box labeled "Miscellany." Inside, amidst trinkets and faded fabrics, he discovered a letter. The handwriting was his grandmother’s, looping and precise. It was addressed to her boss at the local hospital.

He carefully unfolded the paper. It was a draft of a resignation letter. He felt a deep sense of satisfaction as he read his grandmother's words, filled with bold self-assurance, describing her reasons for leaving her post as head nurse to open a soup kitchen. The language was passionate, full of a fierce dedication to helping others.

He sat on a folding chair in the storage unit, a wide grin breaking across his face. He felt strangely connected to his grandmother. He felt as though he understood her, and himself, a little bit better.

Emotion: hopeful

Cluster: Positive / Joy
PC1 (Valence): 3.68 Positive
PC2 (Disposition): -0.58

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.715
la0.549
B0.468
🤩0.452
enthusi0.433
Suppressed:
S-0.603
不-0.401
😞-0.399
worse-0.394
C-0.386