INDEX

Explanations

names of institutions, specific organizations, or notable titles in professional 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

Personensuche

-0.68

rungsseite

-0.65

tvguidetime

-0.63

 tartalomajánló

-0.59

 дописавши

-0.59

 ddelweddau

-0.58

 CallOverrides

-0.57

󠁴

-0.55

 الحره

-0.54

ंदीखरीदारी

-0.54

POSITIVE LOGITS

EndContext

0.49

#+#

0.46

RectangleBorder

0.38

ToRefresh

0.36

Gem

0.33

 roligt

0.30

平成

0.29

stdc

0.28

unes

0.28

XmlAccessorType

0.28

Activations Density 0.141%

names of institutions, specific organizations, or notable titles in professional contexts

No Comments

No Known Activations