INDEX

Explanations

code-related elements, particularly include statements and header files

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

UnusedPrivate

-0.63

InstrumentedTest

-0.61

-------

-0.59

ChildScrollView

-0.57

 ویکی‌پدی

-0.57

AutoField

-0.53

featureID

-0.52

ContentAsync

-0.52

 surla

-0.51

 invokingState

-0.49

POSITIVE LOGITS

 saurait

0.39

Portail

0.37

 censi

0.36

 trône

0.36

 無地

0.36

好评

0.35

 suro

0.35

BeginInit

0.35

ไร

0.35

 Secretaría

0.35

Activations Density 0.008%

code-related elements, particularly include statements and header files

No Comments

No Known Activations