INDEX

Explanations

phrases related to accreditation or official recognition

oai_token-act-pair · gpt-3.5-turbo

references to accreditation and standards in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

OPLE

-0.78

geist

-0.75

NING

-0.72

bian

-0.71

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.68

erness

-0.66

 hope

-0.66

haar

-0.66

çķ

-0.65

 Dare

-0.64

POSITIVE LOGITS

redited

1.21

ompl

1.14

reditation

1.14

ommod

1.13

enture

1.09

laim

1.03

identally

1.02

used

1.02

iona

1.00

idental

0.98

Activations Density 0.007%

phrases related to accreditation or official recognition

references to accreditation and standards in various contexts

No Comments

No Known Activations