INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

CardHeader

-0.95

Бахар

-0.88

+#+

-0.80

 smtplib

-0.80

SupportActionBar

-0.74

inflater

-0.70

ExtendWith

-0.68

 BorderRadius

-0.63

</caption>

-0.63

PhysRevD

-0.62

POSITIVE LOGITS

<0x00>

1.53

PPM

0.75

]--;

0.75

 Lanzarote

0.70

 Nook

0.70

 Yorke

0.70

webElementXpaths

0.68

 XCTest

0.68

lamabad

0.68

iPod

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.