INDEX

Explanations

instances when the writer is discussing personal or work-related updates

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

Reviewer

-0.79

SPONSORED

-0.78

âĢ¢âĢ¢

-0.72

cart

-0.70

Enlarge

-0.66

toggle

-0.66

Critics

-0.66

lect

-0.66

 conformity

-0.64

¿½

-0.64

POSITIVE LOGITS

 hope

1.42

 anticipate

1.34

'll

1.31

 intend

1.29

 apologize

1.28

 apologise

1.24

 promise

1.20

 expect

1.18

 await

1.08

've

1.05

Activations Density 0.259%

instances when the writer is discussing personal or work-related updates

No Comments

No Known Activations