INDEX

Explanations

patterns of tabulation or structured formatting in the document

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

numbers and labels

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



-0.67

Ӕ

-0.62

dafx

-0.60

QMetaType

-0.59

kháu

-0.58

 noDo

-0.57

Aholisi

-0.55

IntOverflow

-0.55

 Infórmanos

-0.54

EndGlobalSection

-0.54

POSITIVE LOGITS

MenuInflater

0.51

 title

0.47

 StatefulWidget

0.43

WEBPACK

0.40

bak

0.40

 Title

0.39

ed

0.39

Future

0.38

 TITLE

0.38

 radio

0.38

Activations Density 0.030%

patterns of tabulation or structured formatting in the document

numbers and labels

No Comments

No Known Activations