INDEX

Explanations

chapter titles containing numbers and colons

oai_token-act-pair · gpt-3.5-turbo

textual references to chapters and sections in a document

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 outraged

-0.73

 franc

-0.71

 appalled

-0.67

 angered

-0.67

kinson

-0.66

 opin

-0.66

 saddened

-0.66

 disgusted

-0.66

 horrified

-0.65

intent

-0.65

POSITIVE LOGITS

 Introduction

0.84

yip

0.74

 Conclusion

0.74

ħĭ

0.67

cation

0.66

 Formation

0.65

£ı

0.65

 Invasion

0.65

 Annual

0.64

å°Ĩ

0.63

Activations Density 0.166%

chapter titles containing numbers and colons

textual references to chapters and sections in a document

No Comments

No Known Activations