INDEX

Explanations

words indicating availability or accessibility of resources

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

QMetaType

-0.62

 Roskov

-0.50

 kapture

-0.49

MessageTagHelper

-0.48

rungsseite

-0.48

 defaultstate

-0.47

\{\\

-0.47

FieldOffsetTable

-0.45

Motto

-0.44



-0.43

POSITIVE LOGITS

 nakalista

0.47

ⓧ

0.46

 viewDidLoad

0.45

IsMutable

0.43

хіво

0.42

 يتيمه

0.42

enumi

0.40

spread

0.40

Thrown

0.38

 forState

0.38

Activations Density 0.051%

words indicating availability or accessibility of resources

No Comments

No Known Activations