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

setVerticalGroup

-0.48

 visiting

-0.43

Revenir

-0.34

sweise

-0.33

 createState

-0.33

 setShow

-0.32

 raising

-0.32

red

-0.31

 floods

-0.31

 police

-0.31

POSITIVE LOGITS

SpringRunner

5.34

 للاسماء

0.83

 GenerationType

0.81

AndroidJUnit

0.81

يكب

0.78

 فريبيس

0.77

SpringBootTest

0.77

ScopeManager

0.76

ArrowToggle

0.72

MigrationBuilder

0.70

Activations Density 0.000%

No Known Activations

This feature has no known activations.