INDEX

Explanations

technical terms and acronyms

oai_token-act-pair · gpt-3.5-turbo

keywords and phrases related to various types of news and reports

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

»Ĵ

-0.58

 millenn

-0.58

 Lanka

-0.57

åŃĲ

-0.56

ranging

-0.54

anmar

-0.53

¿

-0.52

untled

-0.52

drawn

-0.52

rive

-0.52

POSITIVE LOGITS

 exists

1.51

is

1.38

was

1.36

has

1.36

 occurs

1.34

 gets

1.29

 awaits

1.29

 resides

1.29

 goes

1.28

 happens

1.28

Activations Density 0.688%

technical terms and acronyms

keywords and phrases related to various types of news and reports

No Comments

No Known Activations