INDEX

Explanations

instances of the word "street" within a document

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of a specific individual's name or identity

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

 Tornado

-0.71

Mos

-0.65

Miller

-0.64

ç¥ŀ

-0.63

tests

-0.60

STD

-0.60

cano

-0.59

soDeliveryDate

-0.58

 Angels

-0.58

 Spectre

-0.57

POSITIVE LOGITS

reet

1.30

eering

1.07

ing

1.01

ronics

0.93

ype

0.87

rial

0.82

robe

0.81

ypes

0.79

ombo

0.78

reetings

0.78

Activations Density 0.008%

instances of the word "street" within a document

repeated mentions of a specific individual's name or identity

No Comments

No Known Activations