INDEX

Explanations

lists or enumerations of items or attributes

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

`coord` reverse `ng` `h`

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

 autorytatywna

-0.85

 الرياضيه

-0.60

msgTypes

-0.60

LookAnd

-0.59

DebuggerStep

-0.58

الحياه

-0.56

 betweenstory

-0.53

verifyException

-0.52

 ویکی‌پدی

-0.52

afficheront

-0.52

POSITIVE LOGITS

 hObject

0.45

ttino

0.39

 choice

0.39

 spool

0.39

endesk

0.39

 EventArgs

0.39

stood

0.38

 window

0.37

ForRow

0.36

 Brown

0.36

Activations Density 0.009%

lists or enumerations of items or attributes

`coord` reverse `ng` `h`

No Comments

No Known Activations