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).

Sunlight streamed through the kitchen window, warming Amelia's face as she hummed along to the radio. She flipped pancakes with practiced ease, a wide grin stretching across her features. The smell of maple syrup filled the air, mingling with the scent of fresh coffee. This was her perfect Saturday morning โ€“ just her, the music, and a stack of golden-brown deliciousness.

A notification pinged on her phone, and she glanced at the screen. A friend request. From Brad. Brad, who used to delight in tripping her in the school hallway and stealing her lunch money. Amelia's smile faltered, replaced by a momentary frown. It was fleeting, though. She shrugged, grabbed her phone, and tapped 'confirm.' Her day wasn't going to be ruined by a relic from the past.

Emotion: happy

Cluster: Positive / Joy
PC1 (Valence): 3.98 Positive
PC2 (Disposition): -1.14

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.676
excitedly0.533
enthusi0.505
de0.498
๐Ÿคฉ0.495
Suppressed:
S-0.896
๐Ÿ˜ž-0.430
ness-0.410
ใใ‚Œใงใ‚‚-0.387
L-0.379