INDEX

Explanations

terms related to corrections and amendments

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

äfts

-0.53

VersionUID

-0.52

-0.49

readString

-0.48

 חיצוניים

-0.48

 tableName

-0.47

getPackageName

-0.47

PackageName

-0.46

 strlen

-0.46

 isLoading

-0.45

POSITIVE LOGITS

 correction

1.01

 Correction

0.98

correction

0.90

 Corrected

0.88

 corrections

0.88

 Corrections

0.85

 corregir

0.85

 correcting

0.84

 corrects

0.84

 corrected

0.82

Activations Density 0.014%

terms related to corrections and amendments

No Comments

No Known Activations