INDEX

Explanations

terms related to various types of mass events, especially negative or violent ones like mass shootings, mass murder, mass deportation, etc

oai_token-act-pair · gpt-3.5-turbo

references to mass-related events and incidents

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

nos

-0.83

BILITIES

-0.80

thy

-0.78

yssey

-0.75

uana

-0.74

ilk

-0.71

ĪĴ

-0.71

thens

-0.71

izons

-0.71

shire

-0.70

POSITIVE LOGITS

achusetts

1.44

 quantities

0.81

 transit

0.80

 exodus

0.76

 incarceration

0.76

aging

0.75

achus

0.74

 transportation

0.70

 graves

0.69

 istg

0.68

Activations Density 0.016%

terms related to various types of mass events, especially negative or violent ones like mass shootings, mass murder, mass deportation, etc

references to mass-related events and incidents

No Comments

No Known Activations