INDEX

Explanations

functions related to clearing or resetting values

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

WebElementEntity

-0.57

 AttributeSet

-0.48

:✨

-0.48

tagHelper

-0.42

ReusableCell

-0.42

RegistryLite

-0.41

 Eing

-0.41

 Niem

-0.41

 snippetHide

-0.40

 Sala

-0.40

POSITIVE LOGITS

Clear

0.45

Derbyniad

0.44

clear

0.42

 Clear

0.42

PreExecute

0.40

 Numerade

0.39

WriteBarrier

0.39

 삭제

0.38

CLEAR

0.38

 ruptura

0.38

Activations Density 0.010%

functions related to clearing or resetting values

No Comments

No Known Activations