INDEX

Explanations

terms related to diseases and health organizations, specifically mentioning the Centers for Disease Control and Prevention (CDC)

oai_token-act-pair · gpt-3.5-turbo

references to diseases and health-related terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

milo

-0.72

*/(

-0.68

comings

-0.68

dyl

-0.68

skirts

-0.67

verts

-0.67

uckland

-0.67

jug

-0.67

versions

-0.66

eton

-0.65

POSITIVE LOGITS

 Disease

1.13

 Prevention

1.02

 Outbreak

0.99

 Diseases

0.94

 Infect

0.92

 Rating

0.88

 Risk

0.87

 Affairs

0.87

 Disorders

0.85

 Reconstruction

0.84

Activations Density 0.007%

terms related to diseases and health organizations, specifically mentioning the Centers for Disease Control and Prevention (CDC)

references to diseases and health-related terms

No Comments

No Known Activations