INDEX

Explanations

mentions of the country "North Korea."

oai_token-act-pair · gpt-3.5-turbo

references to North Korea

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

TABLE

-0.84

 "$:/

-0.84

NRS

-0.74

ATURE

-0.72

ILA

-0.68

hig

-0.68

PLIC

-0.68

TED

-0.68

###

-0.68

TING

-0.67

POSITIVE LOGITS

 Korea

1.16

 Koreans

1.14

 Korean

0.97

ampton

0.86

uclear

0.85

 Carolina

0.82

 Kurdistan

0.78

 Pole

0.76

 Dakota

0.72

itary

0.70

Activations Density 0.022%

mentions of the country "North Korea."

references to North Korea

No Comments

No Known Activations