INDEX

Explanations

mentions of the Centers for Disease Control and Prevention (CDC)

oai_token-act-pair · gpt-3.5-turbo

references to the CDC (Centers for Disease Control and Prevention)

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

 istg

-0.85

ighters

-0.79

itars

-0.78

wagen

-0.74

estate

-0.69

emale

-0.69

othal

-0.68

othes

-0.65

edom

-0.65

coni

-0.64

POSITIVE LOGITS

CDC

0.98

CDC

0.89

alse

0.74

MD

0.71

VD

0.68

DERR

0.68

 Zika

0.68

 Commissioner

0.67

 Outbreak

0.67

 Infect

0.66

Activations Density 0.018%

mentions of the Centers for Disease Control and Prevention (CDC)

references to the CDC (Centers for Disease Control and Prevention)

No Comments

No Known Activations