INDEX

Explanations

words related to magnification or greatness

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

OPLE

-0.83

thia

-0.73

FAULT

-0.71

lde

-0.70

TAG

-0.70

pta

-0.70

Pg

-0.67

BOOK

-0.66

 Territories

-0.66

 Warriors

-0.66

POSITIVE LOGITS

olia

1.41

ificent

1.30

animous

1.21

ific

1.13

ificant

1.07

itudes

1.05

enaries

0.92

itude

0.91

itud

0.89

ates

0.85

Activations Density 0.029%

words related to magnification or greatness

No Comments

No Known Activations