INDEX

Explanations

references to wings and wing structures

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

Cea

-0.48

 المعيارى

-0.46

ประ

-0.44

!</

-0.43

</caption>

-0.42

Laj

-0.42

 Nagy

-0.42

 ประ

-0.41

testens

-0.41

Ca

-0.41

POSITIVE LOGITS

 Wing

1.25

Wing

1.22

 WING

1.16

 wing

1.15

 Wings

1.12

 wings

1.06

Wings

1.05

wing

1.05

wings

0.98

WING

0.93

Activations Density 0.005%

references to wings and wing structures

No Comments

No Known Activations