INDEX

Explanations

word "cat" or "catastrophic."

oai_token-act-pair · gpt-3.5-turbo

occurrences and references to "cat."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 perme

-0.69

 Vander

-0.67

 enriched

-0.64

unb

-0.64

 steroids

-0.64

 Gree

-0.63

 hemp

-0.63

 HuffPost

-0.60

 Fargo

-0.60

 Morales

-0.60

POSITIVE LOGITS

cat

1.52

aclysm

1.21

cats

1.19

alog

1.18

alogue

1.18

Cat

1.17

alyst

1.14

hedral

1.12

apult

1.07

rina

1.03

Activations Density 0.005%

word "cat" or "catastrophic."

occurrences and references to "cat."

No Comments

No Known Activations