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

inheritdoc

-0.43

cinnati

-0.41

 Soviet

-0.39

ben

-0.38

fun

-0.37

 inter

-0.37

ibatis

-0.36

parti

-0.35

аль

-0.35

dubbo

-0.35

POSITIVE LOGITS

expandindo

0.66

verifyException

0.60

Personendaten

0.60

RegressionTest

0.59

 linkovi

0.58

 étoient

0.57

 propOrder

0.57

 Numerade

0.56

 intptr

0.56

 autorytatywna

0.54

Activations Density 0.000%

No Known Activations

This feature has no known activations.