INDEX

Explanations

references to different parts or sections in a series or a larger entity

oai_token-act-pair · gpt-3.5-turbo

references to parts or sections of works, particularly in serialized content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ydia

-0.74

 sugg

-0.67

owl

-0.66

unning

-0.66

aults

-0.65

faced

-0.63

mails

-0.61

ages

-0.59

 handle

-0.59

aila

-0.59

POSITIVE LOGITS

 Part

3.60

Part

2.49

part

1.96

 Parts

1.86

 PART

1.85

Pt

1.73

 part

1.58

PART

1.54

 Chapter

1.27

 Volume

1.20

Activations Density 0.010%

references to different parts or sections in a series or a larger entity

references to parts or sections of works, particularly in serialized content

No Comments

No Known Activations