INDEX

Explanations

mentions of 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 @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

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

-0.79

llo

-0.78

igue

-0.78

Baltimore

-0.74

MENTS

-0.72

gerald

-0.71

byss

-0.70

Mayor

-0.69

Sensor

-0.68

Arthur

-0.67

POSITIVE LOGITS

orea

1.06

ongyang

0.89

 Pyongyang

0.87

 missile

0.86

 defect

0.84

 missiles

0.82

 Korea

0.81

 blackmail

0.81

 dictator

0.81

 detonated

0.81

Activations Density 0.026%

mentions of the country "North Korea."

mentions of North Korea

No Comments

No Known Activations