INDEX

Explanations

dates and months

oai_token-act-pair · gpt-3.5-turbo

dates and the months associated with them

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

ħĭ

-0.89

Reviewer

-0.85

GoldMagikarp

-0.80

ascript

-0.80

intendent

-0.75

alach

-0.72

trak

-0.72

Interstitial

-0.72

ļéĨĴ

-0.70

awaru

-0.69

POSITIVE LOGITS

nard

0.94

ween

0.87

flower

0.82

 Hare

0.75

0.74

urnal

0.74

oub

0.71

sburg

0.69

Aug

0.69

bug

0.68

Activations Density 0.091%

dates and months

dates and the months associated with them

No Comments

No Known Activations