INDEX

Explanations

words related to clothing items, specifically "blouse."

oai_token-act-pair · gpt-3.5-turbo

references to housing

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

izable

-0.77

ocular

-0.74

 Vaughan

-0.72

imated

-0.72

ellen

-0.71

mie

-0.71

icia

-0.67

OY

-0.66

 Hayward

-0.66

aylor

-0.66

POSITIVE LOGITS

xual

1.34

manship

1.08

ouse

0.96

ktop

0.81

mble

0.78

idon

0.75

uper

0.73

case

0.70

ULTS

0.70

 palms

0.69

Activations Density 0.033%

words related to clothing items, specifically "blouse."

references to housing

No Comments

No Known Activations