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 morning sun streamed through the office window, illuminating dust motes dancing in the air. Sarah adjusted her chair, a slight smile playing on her lips. Today was the day she'd been waiting for: introducing the new hire, Emily, to the ins and outs of her role. A lightness filled her as she thought about Emily thriving, succeeding, and potentially exceeding her own achievements.

Sarah reached for her coffee, the ceramic warm against her fingers. A fresh start. This felt like a fresh start, not just for the company, but for her, too.

"Alright, Emily," Sarah began, her voice brimming with enthusiasm, "Let's dive in. I'm so excited to share what I know."

Emotion: hope

Cluster: Positive / Joy
PC1 (Valence): 3.59 Positive
PC2 (Disposition): -0.12

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de1.065
la0.865
B0.465
🀩0.412
H0.407
Suppressed:
😞-0.459
C-0.451
不-0.449
😣-0.443
S-0.442