INDEX

Explanations

Boolean values and their associated declarations

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

 Hamma

-0.40

 uitge

-0.40

ɕ

-0.40

 behalten

-0.38

roz

-0.38

 without

-0.37

 independiente

-0.36

 Shank

-0.36

 collè

-0.36

ennan

-0.35

POSITIVE LOGITS

 defaultstate

0.70

+#+

0.68

enterOuterAlt

0.67

GEBURTSDATUM

0.65

UnusedPrivate

0.64

tagHelperRunner

0.60

 CreateTagHelper

0.57

setVerticalGroup

0.57

ValueStyle

0.57

OCCURRED

0.56

Activations Density 0.003%

Boolean values and their associated declarations

No Comments

No Known Activations