INDEX

Explanations

code snippets and technical terms

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

HTML/CSS class attributes and code structure markers in programming contexts.

oai_token-act-pair · claude-4-5-sonnet Triggered by @bfgcn7frf5

The neuron is triggered by code identifiers—variable, class, or attribute names (e.g. camelCase or similar tokens) in source snippets.

oai_token-act-pair · o4-mini Triggered by @jyhe0408

New Auto-Interp

Configuration

google/gemma-scope-27b-pt-res/layer_10/width_131k

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 заболевания

-1.44

 Also

-1.34

뚝

-1.32

 Ayrıca

-1.30

appliquer

-1.27

 Envío

-1.26

سجيلات

-1.25

!(

-1.24

ʵ

-1.23

ative

-1.22

POSITIVE LOGITS

vať

1.34

ні

1.25

 boots

1.24

<strong>

1.23

 ہیں

1.22

zijn

1.22

Sunny

1.21

 prospects

1.20

 among

1.17

탉

1.17

Activations Density 0.005%

code snippets and technical terms

HTML/CSS class attributes and code structure markers in programming contexts.

The neuron is triggered by code identifiers—variable, class, or attribute names (e.g. camelCase or similar tokens) in source snippets.

No Comments

No Known Activations

code snippets and technical terms

HTML/CSS class attributes and code structure markers in programming contexts.

The neuron is triggered by code identifiers—variable, class, or attribute names (e.g. camelCase or similar tokens) in source snippets.

No Comments

No Known Activations