INDEX

Explanations

abbreviations or representations related to "ND" which likely stands for "not determined."

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

<bos>

-0.66

JsonPropertyName

-0.59

 localObject

-0.57

存于互联网档案馆

-0.56

parametrize

-0.54

 AttributeSet

-0.54

 >=",

-0.53

 useNavigate

-0.52

googleapis

-0.50

spor

-0.50

POSITIVE LOGITS

nd

0.94

ND

0.88

ND

0.81

nd

0.80

Nd

0.68

Nd

0.65

OND

0.64

NDC

0.64

Ond

0.63

nnd

0.62

Activations Density 0.010%

abbreviations or representations related to "ND" which likely stands for "not determined."

No Comments

No Known Activations