INDEX

Explanations

.net types

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

The neuron is triggered by fully qualified .NET type names—specifically by the “System.…” namespace qualifiers.

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

the

-1.59

彼らの

-1.49

Niveau

-1.49

詳しい

-1.45

 größe

-1.43

 cualquier

-1.42

ביו

-1.40

部分は

-1.39

 tenté

-1.38

Quelle

-1.38

POSITIVE LOGITS

 only

1.63

as

1.56

for

1.56

 because

1.35

1.34

 started

1.34

 with

1.33

 since

1.30

 most

1.30

 Just

1.27

Activations Density 0.079%

.net types

The neuron is triggered by fully qualified .NET type names—specifically by the “System.…” namespace qualifiers.

No Comments

No Known Activations

.net types

The neuron is triggered by fully qualified .NET type names—specifically by the “System.…” namespace qualifiers.

No Comments

No Known Activations

.net types

The neuron is triggered by fully qualified .NET type names—specifically by the “System.…​” namespace qualifiers.

No Comments

No Known Activations

.net types

The neuron is triggered by fully qualified .NET type names—specifically by the “System.…​” namespace qualifiers.

No Comments

No Known Activations

The neuron is triggered by fully qualified .NET type names—specifically by the “System.…” namespace qualifiers.

The neuron is triggered by fully qualified .NET type names—specifically by the “System.…” namespace qualifiers.