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).

"Just another day," Mrs. Evans muttered to herself, staring into the mirror. Her reflection mirrored her mood: the corners of her mouth turned down, her gaze lacking focus. The morning routine felt like a slow, deliberate march towards a day she wasn't sure she could face.

A knock at her door. It was the school secretary with a small, unassuming envelope. No postage, no sender’s address, just her name, neatly printed. She tore it open with a sigh.

Inside, a single pressed flower – a vibrant blue forget-me-not – accompanied a handwritten note. "Remembered you, Ms. Evans. Hope things are going okay." The delicate beauty of the flower arrested her, holding her attention. A small tear traced a path down her cheek.

Emotion: droopy

Cluster: Fatigue / Lethargy
PC1 (Valence): -1.15 Negative
PC2 (Disposition): -1.23

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😞0.303
无法0.252
い぀も0.238
πŸ˜”0.230
:(0.228
Suppressed:
l-0.360
P-0.356
B-0.339
de-0.334
!-0.239