INDEX

Explanations

phrases related to specific actions or events within a process or narrative

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

BuyableInstoreAndOnline

-0.73

but

-0.52

sbm

-0.52

rored

-0.50

ursday

-0.50

Zup

-0.49

stood

-0.49

oris

-0.48

estern

-0.46

endon

-0.45

POSITIVE LOGITS

 afterward

1.12

 afterwards

1.05

 accordingly

1.00

*.

0.98

 instead

0.98

 ðŁĻĤ

0.91

.?

0.91

!.

0.87

.*

0.87

 whenever

0.86

Activations Density 0.947%

phrases related to specific actions or events within a process or narrative

No Comments

No Known Activations