INDEX

Explanations

keywords associated with legal actions and events

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

BeginContext

-0.49

Autoritní

-0.46

IndentedString

-0.44

DebuggerNonUser

-0.43

goc

-0.42

CloseOperation

-0.40

########.

-0.39

PropertyChanging

-0.39

Speaking

-0.39

 Speaking

-0.38

POSITIVE LOGITS

matchCondition

0.47

PARATUS

0.44

 arrests

0.44

 rings

0.39

 confiscation

0.39

blocking

0.38

Arrest

0.37

 arresting

0.37

 trapping

0.36

arrest

0.36

Activations Density 0.156%

keywords associated with legal actions and events

No Comments

No Known Activations