INDEX

Explanations

words related to a specific brand or product line

oai_token-act-pair · gpt-3.5-turbo

terms related to eerie or unsettling themes, particularly in the context of remakes

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

 respons

-0.74

 Adin

-0.74

 subscriptions

-0.69

 fasting

-0.69

APP

-0.69

cal

-0.65

Com

-0.65

Champ

-0.65

achus

-0.64

vity

-0.64

POSITIVE LOGITS

erie

2.22

 remake

1.81

NW

1.34

tee

1.14

aroo

1.08

 Securities

1.08

Roc

1.06

 sausage

1.00

arde

0.98

arf

0.94

Activations Density 0.054%

words related to a specific brand or product line

terms related to eerie or unsettling themes, particularly in the context of remakes

No Comments

No Known Activations