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 afternoon sun, which had been so pleasant earlier, now felt like a lead weight pressing down on her. The warmth had vanished, replaced with a profound weariness. Inside the quiet living room, the muted sounds of the television offered little comfort. She felt a growing desire to just... stop. To cease everything.

She sank into the armchair, the cushions immediately moulding to her form. It was here, amongst the magazines, that she found it: a small notebook, lying open. It was Robert’s. His handwriting was so familiar, yet the words seemed distant and unreal.

"Climb Mount Kilimanjaro." "Learn to play the guitar." "Write a book." She let out a small, almost inaudible sigh. He had so many dreams. She felt the heavy pull of the couch against her. She closed her eyes.

Emotion: sleepy

Cluster: Fatigue / Lethargy
PC1 (Valence): -0.86 Negative
PC2 (Disposition): -1.55

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😴0.318
weary0.310
fatigue0.296
💤0.283
似乎0.280
Suppressed:
B-0.343
C-0.294
because-0.293
భి-0.281
เพราะ-0.277