INDEX

Explanations

various forms of punctuation and formatting elements

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

mathematical notation and code delimiters

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

Jeografia

-0.65

tangentMode

-0.61

fillType

-0.59

rungsseite

-0.56

styleType

-0.56

səhifə

-0.55

OGND

-0.53

alum

-0.53

%"),

-0.52

twimg

-0.52

POSITIVE LOGITS

 AssemblyTitle

0.36

 käs

0.32

 fédé

0.32

beforeEach

0.30

 particulières

0.30

LEGGI

0.30

OuterClass

0.30

 gratuits

0.29

 peine

0.29

 uñas

0.29

Activations Density 0.002%

various forms of punctuation and formatting elements

mathematical notation and code delimiters

No Comments

No Known Activations