INDEX

Explanations

command-related keywords and structures in a programming context

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

command arguments

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_24/width_131k/average_l0_97

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.24.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

 queſta

-0.61

AntiForgeryToken

-0.59

KURZBESCHREIBUNG

-0.57

adpleegd

-0.56

脚注の使い方

-0.55

 المعيارى

-0.54

MigrationBuilder

-0.52

 мәкал

-0.52

$_"

-0.51

 سكانية

-0.50

POSITIVE LOGITS

“

0.52

AddWithValue

0.42

‘

0.39

StackSize

0.37

 thermometer

0.37

RectangleBorder

0.36

0.35

 Thermometer

0.35

 ‘‘

0.34

 thermometers

0.34

Activations Density 0.019%

command-related keywords and structures in a programming context

command arguments

No Comments

No Known Activations