INDEX

Explanations

phrases related to filling gaps or positions

oai_token-act-pair · gpt-3.5-turbo

references to various types of gaps or voids that need to be filled

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

xual

-0.85

schild

-0.74

Magikarp

-0.70

zees

-0.68

nature

-0.66

 Telesc

-0.66

 Kling

-0.66

zai

-0.65

urtles

-0.65

ĪĴ

-0.62

POSITIVE LOGITS

 coffers

1.10

gap

0.87

 cavity

0.76

 remainder

0.74

 vacancy

0.72

 gaps

0.72

 cavern

0.71

 void

0.71

 ranks

0.70

 holes

0.70

Activations Density 0.136%

phrases related to filling gaps or positions

references to various types of gaps or voids that need to be filled

No Comments

No Known Activations