INDEX

Explanations

occurrences of the word "one" and its variations

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

one

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

BibitemShut

-0.54

dbach

-0.50

iastical

-0.49

 rilass

-0.49

romyal

-0.48

 esercit

-0.48

orological

-0.48

uscular

-0.47

 arcilla

-0.47

SharedCtor

-0.46

POSITIVE LOGITS

One

1.22

One

1.13

ONE

1.05

one

1.02

ONE

1.00

getOne

0.80

 oneness

0.80

one

0.78

Ones

0.77

findOne

0.72

Activations Density 0.046%

occurrences of the word "one" and its variations

one

No Comments

No Known Activations