INDEX

Explanations

terms related to data filtering and quality control processes in analyses

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

onAttach

-0.44

 analogue

-0.41

 homolog

-0.41

触

-0.38

何でも

-0.37

+#+

-0.37

Alike

-0.37

TextAppearance

-0.37

}{@

-0.36

 žem

-0.36

POSITIVE LOGITS

RegressionTest

0.57

GEBURTSDATUM

0.51

complexContent

0.51

ंदीखरीदारी

0.51

 resourceCulture

0.51

InputBorder

0.48

 rejection

0.45

Démographie

0.43

 Akismet

0.43

 rechazo

0.42

Activations Density 0.078%

terms related to data filtering and quality control processes in analyses

No Comments

No Known Activations