INDEX

Explanations

references related to significant events or incidents

oai_token-act-pair · gpt-3.5-turbo

references to significant historical events or artifacts

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

uart

-0.39

enegger

-0.36

Ã±o

-0.36

ividual

-0.34

âĵĺ

-0.31

opted

-0.31

 degraded

-0.31

orf

-0.29

anwhile

-0.29

ague

-0.29

POSITIVE LOGITS

natureconservancy

0.59

terness

0.50

etheless

0.47

ĵĺ

0.43

 arrang

0.41

 gobl

0.40

 welf

0.39

 artif

0.37

wcs

0.37

 hous

0.36

Activations Density 7.606%

references related to significant events or incidents

references to significant historical events or artifacts

No Comments

No Known Activations