INDEX

Explanations

rumors or speculations

oai_token-act-pair · gpt-3.5-turbo

words related to rumors or hearsay

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

 Levi

-0.69

 impunity

-0.66

 learning

-0.65

 Cassidy

-0.62

 Chick

-0.60

 autom

-0.59

 Alexandria

-0.59

 Cuomo

-0.58

RET

-0.58

 electronically

-0.58

POSITIVE LOGITS

pled

1.42

ble

1.36

oured

1.27

inating

1.23

inatory

1.15

ination

1.14

mage

1.05

ours

1.05

BLE

1.05

our

1.05

Activations Density 0.022%

rumors or speculations

words related to rumors or hearsay

No Comments

No Known Activations