INDEX

Explanations

instances of the word "Hell" mentioned in the text

oai_token-act-pair · gpt-3.5-turbo

references to the word "Hell."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

Random

-0.69

 sshd

-0.68

Decre

-0.67

Dex

-0.66

 Surveillance

-0.63

 Borders

-0.62

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.61

Gly

-0.60

Recomm

-0.60

York

-0.60

POSITIVE LOGITS

enic

1.22

hound

1.13

ishly

1.11

bent

1.03

fire

1.01

bender

0.98

ish

0.94

ibur

0.93

rieg

0.93

spawn

0.90

Activations Density 0.026%

instances of the word "Hell" mentioned in the text

references to the word "Hell."

No Comments

No Known Activations