INDEX

Explanations

verbs related to giving permission or enabling something

oai_token-act-pair · gpt-3.5-turbo

phrases indicating permission or the ability to do something

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

Sax

-0.76

 Notting

-0.70

é¾įå¥ĳå£«

-0.68

angel

-0.66

_-

-0.64

ÑĢ

-0.63

olate

-0.63

ravis

-0.62

Balt

-0.61

pots

-0.60

POSITIVE LOGITS

SpaceEngineers

0.76

 unrestricted

0.75

ittee

0.75

 sake

0.71

 access

0.70

 flexibility

0.67

 unlimited

0.67

 lapse

0.65

ollower

0.65

umption

0.64

Activations Density 0.238%

verbs related to giving permission or enabling something

phrases indicating permission or the ability to do something

No Comments

No Known Activations