INDEX

Explanations

numerical data, specifically IP addresses and associated formats

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

ip address segments

np_acts-logits-general · gemini-2.5-flash-lite

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

 kaynağından

-0.59

存于互联网档案馆

-0.58

ScopeManager

-0.56

躇

-0.48

TextSpan

-0.48

SharedCtor

-0.47

AutoSize

-0.45

urator

-0.45

 initComponents

-0.44

 Vikipedi

-0.43

POSITIVE LOGITS

Datuak

0.56

ंदीखरीदारी

0.48

DIPSETTING

0.44

 ويكيپيديا

0.44

sidemargin

0.42

 Monfieur

0.41

 CGRectMake

0.40

 hozz

0.39

 defaultstate

0.38

xima

0.38

Activations Density 0.020%

numerical data, specifically IP addresses and associated formats

ip address segments

No Comments

No Known Activations