INDEX

Explanations

specific numerical data points or mathematical expressions

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

numbers and equations

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

aarrggbb

-0.70

Hentet

-0.60

tagext

-0.59

تقاوى

-0.58

脚注の使い方

-0.55

inSlope

-0.55

XmlAccessType

-0.54

UserScript

-0.50

 ModelExpression

-0.50

 <=",

-0.49

POSITIVE LOGITS

Mayo

0.45

ALLY

0.40

Arc

0.38

 läßt

0.38

ignements

0.36

 imprimée

0.36

Magnolia

0.35

GU

0.35

Rory

0.35

ayak

0.35

Activations Density 0.240%

specific numerical data points or mathematical expressions

numbers and equations

No Comments

No Known Activations