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 chipped mug in Ms. Evans' hand trembled slightly as she took a sip of lukewarm coffee. Every morning was the same; a silent, solitary struggle to find a reason to get out of bed. The faces of her colleagues, their effortless camaraderie, felt like another reminder of her isolation. A hollow ache settled into her bones.

A delivery arrived at her door. Inside, a handmade wooden box, intricately carved with images of books and open doors. A note was taped to the lid: "You unlocked something in me, Ms. Evans. Thank you for never giving up on me, even when I gave up on myself. - Marcus Bell." She ran her fingers across the smooth wood, a small smile finally tugging at her lips.

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
😞0.386
ness0.339
缺乏0.326
Suppressed:
de-0.320
(!)-0.286
随后-0.268
😍-0.262
tc-0.255