INDEX

Explanations

instances of mathematical or scientific notations

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

ftagPool

-0.57

WireFormatLite

-0.46

 ujednoznacz

-0.43

رأ

-0.43

 Wilfred

-0.41

muo

-0.41

っぴ

-0.41

 Indirect

-0.40

 poaching

-0.40

mund

-0.39

POSITIVE LOGITS

HtmlAttribute

0.63

StructEnd

0.59

MessageTagHelper

0.57

 estekak

0.48

TagMode

0.47

ViewFeatures

0.47

 Audiodateien

0.44

 poveznice

0.44

HasColumnName

0.42

سطس

0.41

Activations Density 0.001%

instances of mathematical or scientific notations

No Comments

No Known Activations