INDEX

Explanations

No Explanations Found

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

TintMode

-0.59

 ✭✭

-0.58

 betweenstory

-0.58

MessageTagHelper

-0.56

+:+

-0.55

MigrationBuilder

-0.54

featureID

-0.54

bitField

-0.51

 defaultstate

-0.51

AddressBook

-0.48

POSITIVE LOGITS

SourceChecksum

0.85

tagHelperRunner

0.56

Παραπομπές

0.46

Atsauces

0.46

enterOuterAlt

0.45

Kaynakça

0.36

তথ্যসূত্র

0.36

槃

0.33

 Hidup

0.33

ifstream

0.33

Activations Density 0.000%

No Known Activations

This feature has no known activations.