INDEX

Explanations

chapter numbers

oai_token-act-pair · gpt-3.5-turbo

references to specific chapters in texts

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

zzle

-0.75

oggles

-0.73

yg

-0.70

pes

-0.69

berman

-0.68

 blaster

-0.66

enez

-0.66

fitting

-0.66

 fitt

-0.65

ydia

-0.65

POSITIVE LOGITS

ĸļ

0.95

 chapter

0.92

 chapters

0.84

apeake

0.83

chapter

0.81

hound

0.76

 Chapter

0.75

icularly

0.75

 Chapters

0.73

CHAPTER

0.72

Activations Density 0.015%

chapter numbers

references to specific chapters in texts

No Comments

No Known Activations