INDEX

Explanations

text mentioning specific proper names or locations, such as 'Al-Qaeda', 'Iraqi army', 'Edmonton', 'Robredo family'

oai_token-act-pair · gpt-3.5-turbo

references to al-Qaeda and related entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

Jenn

-0.80

TN

-0.75

 Tackle

-0.72

 Ghostbusters

-0.71

esson

-0.69

Purchase

-0.69

GD

-0.68

 Chaser

-0.68

 Pearce

-0.67

SD

-0.65

POSITIVE LOGITS

al

3.61

Al

1.75

Al

1.45

el

1.34

AL

1.32

AQ

1.25

ibn

1.25

ale

1.19

bin

1.17

Alb

1.16

Activations Density 0.023%

text mentioning specific proper names or locations, such as 'Al-Qaeda', 'Iraqi army', 'Edmonton', 'Robredo family'

references to al-Qaeda and related entities

No Comments

No Known Activations