INDEX

Explanations

structured and formatted content, potentially with mathematical or scientific notation

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

list markers like •, *, -

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

SequentialGroup

-0.67

-0.43

WriteLiteral

-0.42

Diweddarwch

-0.41

LogFactory

-0.39

ViewFeatures

-0.38

<?

-0.38

 دیکھیے

-0.38

 يتيمه

-0.37

 noqa

-0.36

POSITIVE LOGITS



0.54

HomeAsUpEnabled

0.48

reportWebVitals

0.46

⇗

0.43

WriteTagHelper

0.43

 PopupWindow

0.42

 otomatig

0.42

__(/*!

0.42

prnewswire

0.41

timbangan

0.40

Activations Density 0.157%

structured and formatted content, potentially with mathematical or scientific notation

list markers like •, *, -

No Comments

No Known Activations