INDEX

Explanations

phrases related to educational institutions and personnel

oai_token-act-pair · gpt-3.5-turbo

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

 Â£

-0.82

ancies

-0.79

Â£

-0.76

 Tories

-0.76

ergus

-0.71

 Nigel

-0.71

colour

-0.70

Jamie

-0.70

 Whilst

-0.69

 whilst

-0.68

POSITIVE LOGITS

 Armenia

1.49

 Azerbai

1.44

 Armenian

1.37

 Azerbaijan

1.36

ÄŁ

1.33

 Kazakhstan

1.24

Ä

1.23

 Kazakh

1.22

 Belarus

1.20

 Jakarta

1.19

Activations Density 1.267%

phrases related to educational institutions and personnel

No Comments

No Known Activations