INDEX

Explanations

references to blogs

oai_token-act-pair · gpt-3.5-turbo

references to blogs

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Scotia

-0.70

arching

-0.67

 Lauder

-0.66

FUL

-0.65

TRUMP

-0.65

asonic

-0.64

CHO

-0.61

IELD

-0.61

BuyableInstoreAndOnline

-0.60

 foil

-0.58

POSITIVE LOGITS

gers

1.39

ger

1.23

ging

1.22

ged

0.93

osphere

0.90

culosis

0.84

gment

0.84

Blog

0.83

 Blog

0.83

bub

0.83

Activations Density 0.008%

references to blogs

references to blogs

No Comments

No Known Activations