INDEX

Explanations

sequences of numerical data

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

numbers and symbols

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

 "..\..\..\

-0.59

astéroïdes

-0.55

WebElementEntity

-0.55

fromnode

-0.49

 "..\..\

-0.48

webElementXpaths

-0.44

 Савезне

-0.44

]=>

-0.44

 ویکی‌پدی

-0.43

хьтан

-0.43

POSITIVE LOGITS

BibitemShut

0.49

addContainerGap

0.47

 تضيفلها

0.45

 fieltro

0.44

elfare

0.43

tifacts

0.42

KommentareTeilen

0.41

uttavia

0.40

ğından

0.39

asitic

0.39

Activations Density 0.250%

sequences of numerical data

numbers and symbols

No Comments

No Known Activations