INDEX

Explanations

prepositions followed by nouns representing specific locations or entities

oai_token-act-pair · gpt-3.5-turbo

words and phrases that indicate locations or settings

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

igure

-0.68

Tickets

-0.66

uploads

-0.66

 respectively

-0.61

catentry

-0.60

ersen

-0.56

iblings

-0.55

 nods

-0.53

 registrations

-0.53

Reports

-0.53

POSITIVE LOGITS

 someone

1.11

 somebody

1.06

any

1.02

 your

1.01

0.97

an

0.97

ANY

0.97

someone

0.95

 anything

0.89

 oneself

0.88

Activations Density 0.599%

prepositions followed by nouns representing specific locations or entities

words and phrases that indicate locations or settings

No Comments

No Known Activations