INDEX

Explanations

mentions of the hotel chain "Hyatt"

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

 reversible

-0.68

IFIED

-0.65

DERR

-0.62

 orally

-0.61

iants

-0.61

 Fired

-0.60

 blow

-0.57

Adds

-0.57

 Nemesis

-0.57

ificantly

-0.56

POSITIVE LOGITS

brids

1.37

giene

1.33

atts

1.22

undai

1.22

brid

1.18

annis

1.16

rule

1.12

pert

1.00

bris

0.95

uman

0.95

Activations Density 0.017%

mentions of the hotel chain "Hyatt"

No Comments

No Known Activations