INDEX

Explanations

references to the country North Korea

oai_token-act-pair · gpt-3.5-turbo

mentions of 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

++++++++++++++++

-0.79

MENTS

-0.76

é¾įåĸļå£«

-0.72

igue

-0.72

Tags

-0.71

llo

-0.70

vol

-0.70

Sah

-0.69

Mouse

-0.69

BET

-0.68

POSITIVE LOGITS

orea

1.00

 defect

0.83

 detonated

0.80

 meltdown

0.79

 Korea

0.78

ese

0.78

 blackmail

0.77

 submarine

0.76

 dictator

0.76

 Koreans

0.74

Activations Density 0.025%

references to the country North Korea

mentions of North Korea

No Comments

No Known Activations