INDEX

Explanations

references to the retail giant "Walmart"

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

INF

-0.78

ager

-0.68

 fracture

-0.68

ãĥķ

-0.67

 trig

-0.66

CN

-0.64

xual

-0.64

ono

-0.63

ange

-0.62

 âī¡

-0.62

POSITIVE LOGITS

Wal

3.84

Wal

2.75

wal

2.17

wal

1.89

 Walmart

1.76

 Walton

1.50

Kro

1.36

 Walters

1.22

 Walter

1.14

 Sher

1.14

Activations Density 0.020%

references to the retail giant "Walmart"

No Comments

No Known Activations