INDEX

Explanations

corrections or errors within text

oai_token-act-pair · gpt-3.5-turbo

corrections or amendments to previously stated information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

uries

-0.80

ãĤ¼ãĤ¦ãĤ¹

-0.78

thood

-0.77

luaj

-0.74

iors

-0.72

ocious

-0.71

æ©

-0.71

çļ

-0.70

soDeliveryDate

-0.69

aceae

-0.68

POSITIVE LOGITS

 typo

1.03

 deleted

1.01

 incorrectly

0.97

 incorrect

0.97

 revised

0.88

 formatting

0.86

*)

0.85

 mistakenly

0.85

 inaccur

0.85

 omitted

0.84

Activations Density 0.321%

corrections or errors within text

corrections or amendments to previously stated information

No Comments

No Known Activations