INDEX

Explanations

references to literary works or novels

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

xon

-0.83

Downloadha

-0.72

rael

-0.71

 impunity

-0.68

 wors

-0.64

cffffcc

-0.61

repair

-0.61

uppet

-0.61

 Admin

-0.60

olid

-0.59

POSITIVE LOGITS

ties

1.34

izations

1.31

isations

1.18

ization

1.16

isation

1.15

istic

1.08

 manuscript

1.06

ists

1.03

istically

1.00

istics

0.97

Activations Density 0.070%

references to literary works or novels

No Comments

No Known Activations