INDEX

Explanations

phrases related to potential consequences or outcomes

oai_token-act-pair · gpt-3.5-turbo

phrases related to health and economic issues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

catentry

-0.78

DragonMagazine

-0.76

\",

-0.71

NetMessage

-0.70

externalActionCode

-0.67

interstitial

-0.65

udeb

-0.64

agic

-0.64

EStreamFrame

-0.64

sol

-0.64

POSITIVE LOGITS

 succeeds

1.26

 fails

1.18

 hadn

1.17

 weren

1.11

 persists

1.08

 survives

1.05

 ever

0.99

 proves

0.96

 decides

0.95

 pans

0.94

Activations Density 0.207%

phrases related to potential consequences or outcomes

phrases related to health and economic issues

No Comments

No Known Activations