INDEX

Explanations

details related to artificial neural network architectures and their configurations

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

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

 EconPapers

-0.45

DockStyle

-0.39

俵

-0.37

Javadoc

-0.36

انجليز

-0.36

</h6>

-0.36

maxY

-0.35

IDES

-0.35

jboss

-0.35

Britannique

-0.34

POSITIVE LOGITS

CPtr

0.64

 simplifié

0.49

WaitGroup

0.43

ArgumentParser

0.43

 незавершена

0.43

 AssemblyCompany

0.42

DotNetBar

0.41

IsMutable

0.40

 Pertama

0.40

 sonne

0.40

Activations Density 0.005%

details related to artificial neural network architectures and their configurations

No Comments

No Known Activations