INDEX

Explanations

the word "Be" followed by a capitalized word, likely focusing on names or titles

oai_token-act-pair · gpt-3.5-turbo

the name "Be" and its variations in different contexts

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

 firsthand

-0.71

 totality

-0.69

 pole

-0.66

ilver

-0.64

 foil

-0.64

 resonance

-0.64

 procedure

-0.64

 lapse

-0.64

 emerge

-0.64

 partName

-0.63

POSITIVE LOGITS

yond

1.16

aum

1.12

hemoth

1.12

acons

1.10

arers

1.06

friends

1.02

ards

1.02

ech

0.99

ijing

0.98

league

0.97

Activations Density 0.062%

the word "Be" followed by a capitalized word, likely focusing on names or titles

the name "Be" and its variations in different contexts

No Comments

No Known Activations