INDEX

Explanations

specific characters or symbols that indicate important elements in text, likely for formatting or structural purposes

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

bootstrapcdn

-0.48

NamedQuery

-0.41

 IFTT

-0.41

ésult

-0.41

 Wikiseite

-0.41

 Chwiliwch

-0.40

 jadx

-0.40

 alamu

-0.40

 فريبيس

-0.40

ِّف

-0.39

POSITIVE LOGITS

transQ

0.62

Personensuche

0.45

 '\\;'

0.44

ChildIndex

0.40

'{@

0.40

negan

0.40

*/),

0.40

antMatchers

0.39

PerformLayout

0.39

MeasureSpec

0.38

Activations Density 0.000%

specific characters or symbols that indicate important elements in text, likely for formatting or structural purposes

No Comments

No Known Activations