INDEX

Explanations

beginning markers or indicators of new sections in text or data

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

 betweenstory

-0.64

writerow

-0.47

transQ

-0.47

########.

-0.46

 <<<<<<<<<<<<<<

-0.46

SaveChangesAsync

-0.43

tonode

-0.43

Hauptartikel

-0.42

 trasparente

-0.42

AlterField

-0.42

POSITIVE LOGITS

 kaarangay

0.63

 Мексичка

0.59

 Taktlose

0.54

Abitanti

0.50

 صوتيه

0.48

SourceChecksum

0.47

IUrlHelper

0.47

ISupport

0.47

 surla

0.45

 Chwiliwch

0.45

Activations Density 0.015%

beginning markers or indicators of new sections in text or data

No Comments

No Known Activations