INDEX

Explanations

the word "huge" at varying levels of intensity

oai_token-act-pair · gpt-3.5-turbo

instances of the word "huge" or its derivatives, indicating discussions about significant importance or scale

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

Reviewed

-0.87

ices

-0.79

pai

-0.79

manship

-0.79

nery

-0.75

cling

-0.74

former

-0.74

apple

-0.71

robe

-0.70

gemony

-0.70

POSITIVE LOGITS

 amounts

1.28

 amount

1.09

 quantities

1.05

 sums

0.97

 chunks

0.95

 swath

0.94

 disparity

0.91

 chunk

0.91

 strides

0.89

 disproportion

0.88

Activations Density 0.053%

the word "huge" at varying levels of intensity

instances of the word "huge" or its derivatives, indicating discussions about significant importance or scale

No Comments

No Known Activations