INDEX

Explanations

Roman numerals and special characters

oai_token-act-pair · gpt-3.5-turbo

symbols or characters that may indicate formatting or code rather than typical language use

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

kens

-0.74

 Jenner

-0.73

creen

-0.68

 Borders

-0.65

 hiking

-0.62

 Disapp

-0.61

 Raleigh

-0.61

 exiting

-0.60

 Chapel

-0.59

 denim

-0.59

POSITIVE LOGITS

ģ

1.52

Ģ

1.48

ī

1.27

ĥ

1.26

Ĥ

1.22

ĩ

1.21

ĭ

1.19

ı

1.16

Ĭ

1.09

Į

1.09

Activations Density 0.008%

Roman numerals and special characters

symbols or characters that may indicate formatting or code rather than typical language use

No Comments

No Known Activations