INDEX

Explanations

HTML headers or tags used for structuring web content

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

start of user input

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

AddTagHelper

-0.52

 Rptr

-0.49

Rptr

-0.49

AutoField

-0.42

Tikang

-0.41

 Clair

-0.40

platte

-0.40

endphp

-0.40

 pets

-0.40

ungszeit

-0.40

POSITIVE LOGITS

IUrlHelper

0.58

ViewFeatures

0.57

DeleteBehavior

0.57

InjectAttribute

0.54

 незавершена

0.54

 autorytatywna

0.51

 BrowserModule

0.50

 noDo

0.48

ConverterFactory

0.47

мәләр

0.47

Activations Density 0.000%

HTML headers or tags used for structuring web content

start of user input

No Comments

No Known Activations