INDEX

Explanations

phrases related to various medical and health conditions

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

jri

-0.82

tnc

-0.74

MpServer

-0.71

soType

-0.70

çīĪ

-0.68

 è£ıç

-0.68

¬¼

-0.67

Reloaded

-0.66

Ore

-0.64

 Sector

-0.64

POSITIVE LOGITS

 tend

1.33

 prefer

1.29

 instinctively

1.12

 behave

1.12

 perceive

1.10

 crave

1.10

 suffer

1.08

 excel

1.07

 thrive

1.07

 cannot

1.07

Activations Density 0.276%

phrases related to various medical and health conditions

No Comments

No Known Activations