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

-0.57

ThroughAttribute

-0.52

 Garrett

-0.44

ContentAlignment

-0.43

配

-0.43

djangoproject

-0.41

Geb

-0.40

 createState

-0.40

CastException

-0.39

 Bres

-0.38

POSITIVE LOGITS

 strany

0.67

 stran

0.55

 ब्रेकडाउन

0.52



0.50

Demografie

0.48

ſelf

0.47

ſelves

0.46

 ujednoznacz

0.44

Referanser

0.42

 strana

0.42

Activations Density 0.000%

No Known Activations

This feature has no known activations.