INDEX

Explanations

phrases indicating different types or meanings of a subject or situation

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

transQ

-0.69

 estekak

-0.69

webElementXpaths

-0.67

writeFieldEnd

-0.63

migrationBuilder

-0.61

脚注の使い方

-0.60

 kasarigan

-0.59

GEBURTSDATUM

-0.59

rawDesc

-0.56

 AssemblyCulture

-0.56

POSITIVE LOGITS

iecie

0.45

UnusedPrivate

0.39

 kond

0.34

who

0.32

})->

0.30

is

0.30

yks

0.30

ondissement

0.29

awaiter

0.29

itä

0.28

Activations Density 0.035%

phrases indicating different types or meanings of a subject or situation

No Comments

No Known Activations