INDEX

Explanations

phrases related to comparisons and modifications in experimental contexts

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

parsedMessage

-0.75

MLLoader

-0.66

Personendaten

-0.65

AISSEE

-0.64

Personensuche

-0.61

webElementXpaths

-0.61

Diweddarwch

-0.61

nodoc

-0.61

TProtocol

-0.59

 wireType

-0.58

POSITIVE LOGITS

ver

0.40

 substituted

0.40

 instead

0.40

 substitute

0.40

 general

0.40

 slightly

0.39

 same

0.39

Zw

0.38

Same

0.37

Replace

0.36

Activations Density 0.149%

phrases related to comparisons and modifications in experimental contexts

No Comments

No Known Activations