INDEX

Explanations

words related to boundaries or limits being surpassed or violated

oai_token-act-pair · gpt-3.5-turbo

concepts related to limits and boundaries

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

atl

-0.82

oxide

-0.81

ointed

-0.73

nan

-0.67

rano

-0.66

ono

-0.63

WD

-0.63

waukee

-0.62

arma

-0.61

war

-0.61

POSITIVE LOGITS

 bounds

1.31

 empt

0.84

izons

0.78

 boundaries

0.78

ourgeois

0.73

ĸļ

0.71

uit

0.70

DragonMagazine

0.68

MpServer

0.68

uits

0.65

Activations Density 0.003%

words related to boundaries or limits being surpassed or violated

concepts related to limits and boundaries

No Comments

No Known Activations