INDEX

Explanations

medical conditions and related terms, such as diseases and treatments

oai_token-act-pair · gpt-3.5-turbo

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

cially

-0.78

erella

-0.69

bilt

-0.68

Else

-0.67

 glim

-0.63

 cents

-0.63

 damning

-0.62

 scrut

-0.62

cipled

-0.61

neau

-0.61

POSITIVE LOGITS

clerosis

0.88

 Orchestra

0.79

tones

0.74

ombat

0.73

rosis

0.73

ences

0.70

vironments

0.70

lords

0.69

 Syndrome

0.69

insula

0.69

Activations Density 0.509%

medical conditions and related terms, such as diseases and treatments

No Comments

No Known Activations