INDEX

Explanations

terms related to virtual phenomena or concepts in graphs

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-res/layer_12/width_1m/average_l0_107

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

1,048,576

Data Type

float32

Hook Name

blocks.12.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

SOD

-0.56

choly

-0.53

 despedida

-0.50

 capelli

-0.50

Tha

-0.49

putAll

-0.49

 ricordi

-0.49

-0.48

 vivement

-0.48

 contiennent

-0.48

POSITIVE LOGITS

 AssemblyCulture

0.94

webElementXpaths

0.90

djangoproject

0.90

IntoConstraints

0.89

 unknownFields

0.89

VersionUID

0.87

ViewFeatures

0.87

EDEFAULT

0.85

 оригіналу

0.84

Portale

0.84

Activations Density 0.002%

terms related to virtual phenomena or concepts in graphs

No Comments

No Known Activations