INDEX

Explanations

phrases that list categories or options

oai_token-act-pair · gpt-3.5-turbo

phrases indicating the classification or categorization of concepts and reasons

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

ergy

-0.74

Beast

-0.72

Ire

-0.71

bats

-0.71

 Andromeda

-0.68

uddin

-0.66

ebook

-0.64

Downloadha

-0.62

istant

-0.62

lator

-0.59

POSITIVE LOGITS

viz

1.08

%:

1.02

:-

1.00

):

0.85

 simultaneously

0.85

 namely

0.83

0.82

:(

0.80

:#

0.78

:"

0.75

Activations Density 0.132%

phrases that list categories or options

phrases indicating the classification or categorization of concepts and reasons

No Comments

No Known Activations