INDEX

Explanations

terms related to the concept of whitewashing

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "new."

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

 cort

-0.73

 suspic

-0.71

Downloadha

-0.69

 unarmed

-0.67

 politics

-0.66

 legislatures

-0.63

 distracted

-0.63

 politically

-0.62

 falsely

-0.62

 thieves

-0.61

POSITIVE LOGITS

estern

1.13

een

1.09

esome

1.09

riter

1.09

alker

1.00

esley

1.00

ritten

0.93

arp

0.92

ild

0.91

elcome

0.91

Activations Density 0.009%

terms related to the concept of whitewashing

occurrences of the word "new."

No Comments

No Known Activations