INDEX

Explanations

email addresses within text

oai_token-act-pair · gpt-3.5-turbo

instances of email-related contact information

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

Ń·

-1.21

İĭ

-1.02

ĪĴ

-0.96

Ļ½

-0.87

ĸļ

-0.83

¿½

-0.81

©¶æ¥µ

-0.80

±

-0.79

»Ĵ

-0.78

ãĥ¼ãĥĨ

-0.78

POSITIVE LOGITS

âĨ

0.71

tracks

0.64

Cour

0.63

report

0.63

ourt

0.62

leader

0.59

 Panc

0.59

 Gathering

0.58

=>

0.58

},"

0.57

Activations Density 0.060%

email addresses within text

instances of email-related contact information

No Comments

No Known Activations