INDEX

Explanations

mentions of specific names, particularly "Stafford" and "Caldwell"

oai_token-act-pair · gpt-3.5-turbo

references to the name "Stafford" and related entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

Magikarp

-0.91

ntil

-0.75

OHN

-0.75

ocard

-0.73

-0.72

ablishment

-0.71

ccording

-0.70

ideon

-0.70

yip

-0.68

 Gleaming

-0.66

POSITIVE LOGITS

shire

1.62

 Stafford

1.33

ington

0.89

sburgh

0.76

inois

0.75

stown

0.73

iod

0.71

andowski

0.70

aday

0.70

ick

0.69

Activations Density 0.008%

mentions of specific names, particularly "Stafford" and "Caldwell"

references to the name "Stafford" and related entities

No Comments

No Known Activations