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).

"We apologize for the continued delay," the loudspeaker droned, and for once, the words didn't sting. Robert, on his way to a crucial pitch for a major business deal, found himself in a state of controlled anticipation.

He sat down, closed his eyes, and began to brainstorm. He envisioned the deal, the words of his presentation, the perfect timing. But something was different; he was not driven by anxiety.

Instead, he was able to envision the success, with clarity and focus. He felt an unusual sense of calm, a powerful confidence in his ability. He opened his laptop and started refining his pitch, the words flowing effortlessly, fueled by a newfound zeal.

Emotion: inspired

Cluster: Positive / Joy
PC1 (Valence): 2.98 Positive
PC2 (Disposition): 0.04

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
B0.563
invigor0.465
newfound0.446
هذ0.414
используя0.406
Suppressed:
😞-0.481
不-0.431
worse-0.428
😣-0.400
fact-0.389