INDEX

Explanations

phrases related to betrayal or serious crimes

oai_token-act-pair · gpt-3.5-turbo

terms related to treason and serious crimes

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

illon

-0.64

 Etsy

-0.63

bye

-0.61

aternity

-0.61

addle

-0.60

WD

-0.59

Dee

-0.59

 Employees

-0.58

 Lori

-0.58

 Dakota

-0.58

POSITIVE LOGITS

ous

2.11

ously

1.84

OUS

1.05

istically

1.01

istical

1.00

inates

0.96

uously

0.94

istics

0.92

aries

0.91

sts

0.90

Activations Density 0.038%

phrases related to betrayal or serious crimes

terms related to treason and serious crimes

No Comments

No Known Activations