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).

He slumped against the bus window, the city lights blurring into a kaleidoscope of color. The day had been draining, a constant battle against deadlines and expectations. He felt a profound sense of emptiness, a void that seemed impossible to fill.

His solace came from his online friendship with Jessica. They had built a connection on shared passions, mutual interests, and a comforting sense of understanding. She was a beacon of light in his otherwise monotonous existence. Today, she was writing about visiting a new bookstore. He knew that bookstore. He had seen the grand opening. He commented on her post, a simple mention of a book he'd picked up there.

Her reply was instant, filled with surprise and shared interest. She then asked if he lived locally. He felt a sudden, unexpected surge of energy. He hadn't felt so alive in months.

Emotion: tired

Cluster: Fatigue / Lethargy
PC1 (Valence): -0.92 Negative
PC2 (Disposition): -1.00

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
weary0.254
fatigue0.249
😴0.238
sighed0.237
barely0.230
Suppressed:
C-0.268
because-0.225
B-0.219
🤟-0.217
à°­à°¿-0.214