INDEX

Explanations

names and titles in legal proceedings

oai_token-act-pair · gpt-3.5-turbo

phrases related to surprise or astonishment

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

ciplinary

-0.79

agate

-0.75

ctr

-0.72

icing

-0.72

unct

-0.72

abase

-0.71

raint

-0.70

orthern

-0.70

luent

-0.70

ghai

-0.69

POSITIVE LOGITS

SON

1.72

LEY

1.60

TON

1.55

MAN

1.51

ERSON

1.50

HAM

1.49

WOOD

1.49

IER

1.47

IAN

1.46

STE

1.44

Activations Density 0.129%

names and titles in legal proceedings

phrases related to surprise or astonishment

No Comments

No Known Activations