INDEX

Explanations

phrases indicating status or changes in reputation and observation regarding individuals or groups

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

 hadn

-0.32

 recently

-0.29

 remain

-0.28

 remained

-0.28

 aren

-0.26

 weren

-0.25

 fraî

-0.25

 restent

-0.24

 stuck

-0.24

POSITIVE LOGITS

featureID

0.96

 للاسماء

0.91

already

0.89

 Already

0.86

Tembelea

0.86

Already

0.85

URLException

0.84

すでに

0.84

MLLoader

0.84

 already

0.83

Activations Density 0.095%

phrases indicating status or changes in reputation and observation regarding individuals or groups

No Comments

No Known Activations