INDEX

Explanations

capital letters that are not common in regular words

oai_token-act-pair · gpt-3.5-turbo

abbreviations or acronyms, particularly those involving significant names or concepts in a technical or cultural context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 symp

-0.76

Wid

-0.75

 Sally

-0.75

Mum

-0.74

 Nichols

-0.72

 Whites

-0.72

 Percentage

-0.70

 landsl

-0.69

 Nich

-0.68

 Nina

-0.68

POSITIVE LOGITS

ar

1.64

AR

1.48

ars

1.44

arn

1.24

ared

1.24

arp

1.21

ARS

1.20

arat

1.18

art

1.16

aro

1.16

Activations Density 0.193%

capital letters that are not common in regular words

abbreviations or acronyms, particularly those involving significant names or concepts in a technical or cultural context

No Comments

No Known Activations