INDEX

Explanations

keywords related to being hacked or compromised

oai_token-act-pair · gpt-3.5-turbo

words related to the act of hitting or striking

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

olith

-0.75

ieves

-0.72

 Debor

-0.69

âĿ

-0.64

 livest

-0.64

hetically

-0.60

Cla

-0.59

geries

-0.59

psons

-0.59

erva

-0.59

POSITIVE LOGITS

acked

1.32

ACK

1.14

acks

1.07

acking

0.99

acker

0.94

ansas

0.80

ack

0.79

ernel

0.79

enzie

0.78

Accessory

0.77

Activations Density 0.011%

keywords related to being hacked or compromised

words related to the act of hitting or striking

No Comments

No Known Activations