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

ThroughAttribute

-0.68

⤹

-0.56

 trib

-0.53

 ✭✭

-0.52

 jsPsych

-0.52

 bacilli

-0.52

<?

-0.52

 papy

-0.50

 methodName

-0.49



-0.49

POSITIVE LOGITS

Datuak

0.81

UPDATED

0.71

Updated

0.66

updated

0.60

 updated

0.59

帖最后由

0.57

Aktualisiert

0.56

 Updated

0.55

 actualizado

0.55

 updating

0.52

Activations Density 0.000%

No Known Activations

This feature has no known activations.