INDEX

Explanations

phrases related to grouping or categorizing items

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

 Taktlose

-0.47

CFO

-0.46

WLAN

-0.45

 OLED

-0.44

userdata

-0.44

VIAF

-0.44

ЙТЕ

-0.44

UserContext

-0.43

OGND

-0.43

ymo

-0.42

POSITIVE LOGITS

enumi

0.69

 each

0.53

 whichever

0.52

Each

0.49

 particular

0.48

 Ogni

0.47

 Each

0.47

AndEndTag

0.47

 CreateTagHelper

0.47

 jede

0.46

Activations Density 0.047%

phrases related to grouping or categorizing items

No Comments

No Known Activations