INDEX

Explanations

keywords related to errors or mistakes

oai_token-act-pair · gpt-3.5-turbo

instances of the word "err."

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

eph

-0.78

creen

-0.76

xual

-0.76

esville

-0.72

ĺħ

-0.69

rooms

-0.66

Sing

-0.65

nai

-0.65

cape

-0.63

clud

-0.63

POSITIVE LOGITS

untled

1.09

anger

0.98

rr

0.95

idge

0.95

antly

0.93

ands

0.88

andom

0.87

ific

0.86

ange

0.86

uti

0.84

Activations Density 0.015%

keywords related to errors or mistakes

instances of the word "err."

No Comments

No Known Activations