INDEX

Explanations

proper nouns, specifically title case words

oai_token-act-pair · gpt-3.5-turbo

the word "The" as a key indicator within the text

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

alone

-0.72

 beforehand

-0.71

/"

-0.68

 himself

-0.67

 themselves

-0.67

 theirs

-0.66

 afterwards

-0.64

\<

-0.64

gpu

-0.64

 with

-0.62

POSITIVE LOGITS

resa

1.39

oret

1.28

odore

1.23

ories

1.10

atre

1.06

 Latest

0.98

 latest

0.96

 simplest

0.95

 Basics

0.95

orem

0.94

Activations Density 0.315%

proper nouns, specifically title case words

the word "The" as a key indicator within the text

No Comments

No Known Activations