INDEX

Explanations

light and related concepts

np_acts-logits-general · gemini-2.5-flash-lite

the neuron detects section headings / numbered step labels and other high-level discourse markers that start or emphasize explanatory sections.

oai_token-act-pair · gpt-5-mini Triggered by @yooniel31

concepts and discussions related to light and optics, including illumination, color, wavelengths, scattering, and physics-of-light phenomena.

oai_token-act-pair · gpt-5 Triggered by @yooniel31

New Auto-Interp

Configuration

google/gemma-scope-2-27b-it/resid_post_all/layer_41_width_262k_l0_big

Prompts (Dashboard)

238,145 prompts, 512 tokens each

Dataset (Dashboard)

lmsys + oasst1

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 সরবরা

0.42

ுற

0.41

 agricultura

0.40

 gecko

0.39

द्योग

0.38

فور

0.38

 கும்ப

0.38

 volcano

0.38

爺

0.38

នា

0.38

POSITIVE LOGITS

 light

2.13

光

1.83

 Light

1.82

Light

1.81

빛

1.80

light

1.74

 свет

1.71

 lumière

1.69

 cahaya

1.67

luz

1.65

Activations Density 0.018%

light and related concepts

the neuron detects section headings / numbered step labels and other high-level discourse markers that start or emphasize explanatory sections.

concepts and discussions related to light and optics, including illumination, color, wavelengths, scattering, and physics-of-light phenomena.

No Comments

No Known Activations

light and related concepts

the neuron detects section headings / numbered step labels and other high-level discourse markers that start or emphasize explanatory sections.

concepts and discussions related to light and optics, including illumination, color, wavelengths, scattering, and physics-of-light phenomena.

No Comments

No Known Activations