INDEX

Explanations

statistical terms and phrases related to research conclusions and analysis

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

istoitu

-0.75

 EconPapers

-0.73

 perchè

-0.67

رشف

-0.60

 poichè

-0.59

!)

-0.58

 Schluß

-0.57

 ligiloj

-0.57

Geplaatst

-0.57

 שוליים

-0.56

POSITIVE LOGITS

 Afterward

0.54

 afterward

0.49

0.41

0.40

https

0.39

±

0.38

 continúas

0.38

uxxxx

0.37

nullptr

0.34

 nullptr

0.34

Activations Density 0.136%

statistical terms and phrases related to research conclusions and analysis

No Comments

No Known Activations