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 gentle rain felt refreshing on Daniel's face as he walked toward the spot where he had planted the cherry tree. He felt a lightness in his step, like a great weight had been lifted. The tree represented a promise, a future he was eager to enjoy.

He stopped, his smile faltering. The tree was gone. All that remained was a freshly cut stump. A quick flash of anger and disappointment, but it quickly dissolved. He took a deep breath.

"Oh, well," he said aloud, already planning a new venture. "Time to plant something else, I suppose." He ran his hand along the smooth wood, and pulled out his phone to order a new set of seeds.

Emotion: optimistic

Cluster: Positive / Joy
PC1 (Valence): 4.44 Positive
PC2 (Disposition): -0.49

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.856
la0.688
!0.601
enthusi0.530
๐Ÿคฉ0.515
Suppressed:
S-0.891
๐Ÿ˜ž-0.504
L-0.486
ness-0.478
ใใ‚Œใงใ‚‚-0.432