INDEX

Explanations

elements related to programming syntax and structure, particularly in code

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

 itſelf

-0.41

NewLabel

-0.41

wiſe

-0.37

 Majefty

-0.36

XmlRootElement

-0.36

 škoda

-0.35

WriteLiteral

-0.35

ningss

-0.34

PostMapping

-0.33

 stronę

-0.32

POSITIVE LOGITS

SharedDtor

0.68

rungsseite

0.59

__':

0.57

__":

0.54

::*;

0.52

ыгана

0.52

 EconPapers

0.51

belea

0.50

initComponents

0.48

complexContent

0.48

Activations Density 0.007%

elements related to programming syntax and structure, particularly in code

No Comments

No Known Activations