INDEX

Explanations

words related to medication or medical treatments

oai_token-act-pair · gpt-3.5-turbo

words related to comedy or humorous content

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

atchewan

-0.79

accompan

-0.70

İĭ

-0.69

ific

-0.69

animous

-0.69

falls

-0.67

ippi

-0.67

hematically

-0.67

ership

-0.66

arers

-0.66

POSITIVE LOGITS

edy

0.95

LINE

0.80

deen

0.73

ge

0.73

geoning

0.70

lig

0.70

dy

0.68

der

0.66

zzy

0.65

isal

0.64

Activations Density 0.014%

words related to medication or medical treatments

words related to comedy or humorous content

No Comments

No Known Activations