INDEX

Explanations

mentions of the Pentagon

oai_token-act-pair · gpt-3.5-turbo

frequent mentions of the term "Pentagon."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

lihood

-0.76

Interstitial

-0.73

à

-0.72

bars

-0.67

Jenn

-0.67

cause

-0.65

Torrent

-0.65

 toget

-0.65

hirt

-0.64

 constitu

-0.64

POSITIVE LOGITS

 Papers

1.32

 Pentagon

0.93

 Office

0.92

 SEAL

0.91

 contractors

0.84

 Department

0.82

 contractor

0.81

 brass

0.79

 planners

0.78

 inspector

0.78

Activations Density 0.023%

mentions of the Pentagon

frequent mentions of the term "Pentagon."

No Comments

No Known Activations