INDEX

Explanations

updates or announcements within a document

oai_token-act-pair · gpt-3.5-turbo

update notifications or announcements

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

aden

-0.91

athered

-0.79

anship

-0.79

gravity

-0.78

wegian

-0.74

enic

-0.74

uman

-0.73

vasive

-0.70

quartered

-0.68

atom

-0.68

POSITIVE LOGITS

 Update

1.04

 UPDATE

1.00

UPDATE

0.95

:]

0.92

Update

0.90

 Sources

0.83

 EDIT

0.81

!:

0.80

III

0.79

Activations Density 0.035%

updates or announcements within a document

update notifications or announcements

No Comments

No Known Activations