INDEX

Explanations

references to specific mathematical or algorithmic concepts in a structured format

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

 gyhoeddwyd

-0.52

MigrationBuilder

-0.50

 surprised

-0.45

surprised

-0.45

Fot

-0.43

oa̍t

-0.43

addPreferredGap

-0.43

abs

-0.43

ToLeft

-0.42

caps

-0.42

POSITIVE LOGITS

parsedMessage

0.50

liothèque

0.40

Tikang

0.40

 ujednoznacz

0.39

eleste

0.38

stylers

0.37

 OMITBAD

0.36

 financière

0.34

EndContext

0.34

 wzglę

0.34

Activations Density 0.172%

references to specific mathematical or algorithmic concepts in a structured format

No Comments

No Known Activations