INDEX

Explanations

information related to incidents and legal actions

oai_token-act-pair · gpt-3.5-turbo

phrases related to health emergencies or medical crises

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

neapolis

-0.69

bernatorial

-0.66

 Organizations

-0.64

 Theater

-0.64

 Centers

-0.64

uh

-0.64

MFT

-0.64

BYU

-0.63

 Negro

-0.63

 Whitman

-0.63

POSITIVE LOGITS

 organisers

0.98

".

0.92

 organisation

0.91

 recognise

0.90

 analyse

0.89

''.

0.88

 programme

0.86

 licence

0.86

 apologise

0.85

 apologised

0.85

Activations Density 1.758%

information related to incidents and legal actions

phrases related to health emergencies or medical crises

No Comments

No Known Activations