INDEX

Explanations

the presence of coding structure elements and keywords related to data models and API functionality

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

abstract classes and data types

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

CppMethod

-0.52

 kasarigan

-0.49

featureID

-0.49

 lenker

-0.44

IsMutable

-0.42

gsSP

-0.42

SourceChecksum

-0.39

Jeografia

-0.38

setVerticalGroup

-0.38

 jScrollPane

-0.37

POSITIVE LOGITS

WriteTagHelper

0.51

 ویکی‌پدیا

0.50

 henkilö

0.50

Biografia

0.46

сылкі

0.46

 człowie

0.45

vrolet

0.44

IZONA

0.42

 magna

0.41

instancetype

0.41

Activations Density 0.002%

the presence of coding structure elements and keywords related to data models and API functionality

abstract classes and data types

No Comments

No Known Activations