INDEX

Explanations

references to visual collections, catalogs, or repositories

oai_token-act-pair · gpt-4o-mini Triggered by @bot

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 autorytatywna

-0.43

 BoxFit

-0.40

 lệ

-0.37

 Administrativna

-0.35

makeText

-0.35

atguigu

-0.35

 ब्रेकडाउन

-0.35

ãi

-0.34

urb

-0.34

 Petter

-0.34

POSITIVE LOGITS

 &___

0.63

 betweenstory

0.52

verwijspagina

0.51

ChildScrollView

0.50

 OkHttpClient

0.49

TabbedPane

0.48

WireFormatLite

0.47

CodedInputStream

0.46

protoimpl

0.45

 CommonModule

0.45

Activations Density 0.089%

references to visual collections, catalogs, or repositories

No Comments

No Known Activations