INDEX

Explanations

HTML tags and formatted text elements

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

paragraph opening tag

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

 betweenstory

-0.57

новниш

-0.56

 Italijani

-0.54

AutoField

-0.53

#+#

-0.52

+#+

-0.50

Organisateur

-0.50

webElementXpaths

-0.50

 JpaRepository

-0.49

kosh

-0.49

POSITIVE LOGITS

enumi

0.43

Gum

0.40

DDI

0.40

\{\\

0.40

AutoresizingMask

0.38

'{@

0.38

ChildScrollView

0.38

astify

0.37

bigskip

0.37

medskip

0.37

Activations Density 0.173%

HTML tags and formatted text elements

paragraph opening tag

No Comments

No Known Activations