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).

Rain lashed against the windows of the old Victorian house, mirroring the storm inside Leo. He hadn't touched the tea his sister, Clara, made hours ago; the ceramic cup sat, forgotten, on the cluttered kitchen table. Dust motes danced in the lone shaft of sunlight that pierced the gloom, each one a tiny, insignificant speck.

Clara was upstairs, apparently sorting through their grandmother's belongings. He'd offered to help, but she'd waved him away, her voice gentle but firm. "You take a break, Leo. You deserve it." He hadn't bothered to argue. He didnโ€™t deserve anything.

He picked at a loose thread on his worn-out jeans. The silence was deafening, broken only by the rhythmic tick of the grandfather clock in the hallway, each tick a mocking reminder of the passing time he seemed incapable of filling. He felt a hollowness in his chest, a deep ache that no amount of tea or sunshine could soothe.

Emotion: worthless

Cluster: Sadness / Despair
PC1 (Valence): -2.10 Negative
PC2 (Disposition): -0.15

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
L0.427
S0.405
๐Ÿ˜ž0.386
ness0.339
็ผบไน0.326
Suppressed:
de-0.320
(!)-0.286
้šๅŽ-0.268
๐Ÿ˜-0.262
tc-0.255