INDEX

Explanations

references to numerical data related to various categories or subjects

oai_token-act-pair · gpt-3.5-turbo

phrases related to statistical results or numerical data summaries

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

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

 trave

-0.62

 Safari

-0.58

å£«

-0.56

 adhere

-0.56

Progress

-0.55

 Presence

-0.55

 realization

-0.54

 metab

-0.52

 reacts

-0.52

 demonstration

-0.52

POSITIVE LOGITS

 eight

1.26

 nine

1.26

 thirteen

1.24

 seven

1.24

 fourteen

1.23

six

1.22

 eleven

1.21

 sixteen

1.20

 nineteen

1.18

 seventeen

1.17

Activations Density 0.070%

references to numerical data related to various categories or subjects

phrases related to statistical results or numerical data summaries

No Comments

No Known Activations