INDEX

Explanations

phrases related to prohibition or restriction

oai_token-act-pair · gpt-3.5-turbo

references to types or categories of things

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

NAS

-0.75

 VIDEOS

-0.70

UNCH

-0.69

borough

-0.68

ANS

-0.68

 Minutes

-0.63

sbm

-0.63

ITNESS

-0.63

IVER

-0.62

eor

-0.62

POSITIVE LOGITS

ilege

0.79

unia

0.74

lier

0.72

 prov

0.70

 imaginable

0.69

wcsstore

0.65

liest

0.63

 reconciliation

0.63

lihood

0.63

gery

0.63

Activations Density 0.032%

phrases related to prohibition or restriction

references to types or categories of things

No Comments

No Known Activations