INDEX

Explanations

variable assignment keywords

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

code-like syntax in configuration files and programming contexts, particularly where values are being inherited or matched.

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

The neuron detects programming/configuration keywords and property names (e.g. “escape,” “inherited,” “match_parent,” “only,” etc.) across code 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

 then

-2.05

 likely

-1.72

 because

-1.65

 definitely

-1.53

you

-1.52

-1.49

 finally

-1.47

 should

-1.46

 probably

-1.46

صبح

-1.34

POSITIVE LOGITS

1.52

 schuldig

1.50

 Vikipedi

1.46

 июня

1.44

 maravill

1.44

 Григо

1.43

นด์

1.41

 décontract

1.41

ueves

1.41

atste

1.41

Activations Density 0.026%

variable assignment keywords

code-like syntax in configuration files and programming contexts, particularly where values are being inherited or matched.

The neuron detects programming/configuration keywords and property names (e.g. “escape,” “inherited,” “match_parent,” “only,” etc.) across code snippets.

No Comments

No Known Activations

variable assignment keywords

code-like syntax in configuration files and programming contexts, particularly where values are being inherited or matched.

The neuron detects programming/configuration keywords and property names (e.g. “escape,” “inherited,” “match_parent,” “only,” etc.) across code snippets.

No Comments

No Known Activations