INDEX

Explanations

specific numerical values that are considered significant or noteworthy

oai_token-act-pair · gpt-3.5-turbo

special characters or encoding issues within the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 hitch

-0.78

glers

-0.74

 confidentiality

-0.73

nect

-0.73

anooga

-0.72

*/(

-0.70

essential

-0.70

icipated

-0.68

sonian

-0.68

ibles

-0.68

POSITIVE LOGITS

¹

1.24

³

1.13

²

1.12

¸

1.03

Į

1.00

´

0.97

ħ

0.93

ł

0.90

ÙĦ

0.89

°

0.88

Activations Density 0.004%

specific numerical values that are considered significant or noteworthy

special characters or encoding issues within the text

No Comments

No Known Activations