INDEX

Explanations

phrases related to violations of laws or regulations

oai_token-act-pair · gpt-3.5-turbo

references to legal violations and laws

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

isSpecialOrderable

-0.81

soDeliveryDate

-0.78

uilding

-0.74

ussen

-0.72

akeru

-0.71

idates

-0.70

INGTON

-0.70

osaurus

-0.68

acs

-0.68

hin

-0.68

POSITIVE LOGITS

 norms

1.06

 norm

0.95

 principle

0.92

 principles

0.90

 jurisdiction

0.89

 jurisd

0.87

 confidentiality

0.86

 tenets

0.85

 precept

0.84

 sensibilities

0.79

Activations Density 0.351%

phrases related to violations of laws or regulations

references to legal violations and laws

No Comments

No Known Activations