INDEX

Explanations

words related to organization names or entities

oai_token-act-pair · gpt-3.5-turbo

abbreviations and acronyms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Liberties

-0.70

 Herm

-0.64

Vag

-0.63

 Virus

-0.63

 nausea

-0.62

=-=-=-=-

-0.61

 Johns

-0.60

 Lara

-0.60

 pretext

-0.60

 Citizenship

-0.59

POSITIVE LOGITS

ITION

1.36

ITS

1.32

ISH

1.32

FORM

1.27

WORK

1.27

ANT

1.26

IUM

1.25

NER

1.24

REC

1.22

LET

1.21

Activations Density 0.186%

words related to organization names or entities

abbreviations and acronyms

No Comments

No Known Activations