INDEX

Explanations

abbreviations with the letters "S" and "9"

oai_token-act-pair · gpt-3.5-turbo

acronyms and their associated meanings

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

CPR

-0.72

ãĥ£

-0.68

ãĤ¡

-0.68

puter

-0.67

 morphine

-0.67

 unrestricted

-0.64

 bron

-0.63

catentry

-0.63

 stret

-0.63

wcs

-0.62

POSITIVE LOGITS

ESSION

1.21

ustain

1.09

ELF

1.02

eeds

1.02

OUND

1.02

CE

1.01

RS

1.00

essions

0.99

ustainable

0.99

AND

0.98

Activations Density 0.037%

abbreviations with the letters "S" and "9"

acronyms and their associated meanings

No Comments

No Known Activations