INDEX

Explanations

the word "most" in various contexts

oai_token-act-pair · gpt-3.5-turbo

references to "most" in relation to various scenarios or subjects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.7.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

rompt

-0.94

icer

-0.72

orld

-0.72

heid

-0.69

instead

-0.65

moil

-0.64

apist

-0.64

isson

-0.64

nton

-0.63

alid

-0.63

POSITIVE LOGITS

 importantly

0.94

 mornings

0.78

body

0.75

afa

0.75

ONE

0.74

 sane

0.73

 Americans

0.69

THING

0.68

 observers

0.68

 cases

0.67

Activations Density 0.051%

the word "most" in various contexts

references to "most" in relation to various scenarios or subjects

No Comments

No Known Activations