INDEX

Explanations

markers and symbols used in technical or programming 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

 collègues

-0.40

 content

-0.37

 successful

-0.37

content

-0.36

 official

-0.35

 effective

-0.34

 inhoud

-0.34

 nieruch

-0.34

 conteúdos

-0.34

 formal

-0.33

POSITIVE LOGITS

fromnode

0.69

findpost

0.64

RenderAtEndOf

0.59

InstrumentedTest

0.58

oneofs

0.56

makeConstraints

0.56

ContentAsync

0.54

setVerticalGroup

0.52

IndentedString

0.51

AnchorStyles

0.50

Activations Density 0.021%

markers and symbols used in technical or programming contexts

No Comments

No Known Activations