INDEX

Explanations

phrases related to instructions or guidelines

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

 unden

-0.45

 reluct

-0.44

ccording

-0.40

©¶æ¥µ

-0.40

ãĥ´

-0.40

emale

-0.40

mostly

-0.40

vez

-0.39

rer

-0.39

ãĤ©

-0.39

POSITIVE LOGITS

or

2.07

nor

1.62

OR

1.25

Or

1.21

Or

1.19

 either

1.08

or

1.08

nor

1.07

 whatsoever

1.03

 anymore

1.02

Activations Density 1.664%

phrases related to instructions or guidelines

No Comments

No Known Activations