INDEX

Explanations

phrases referring to certain characteristics or types mentioned in a comparison

oai_token-act-pair · gpt-3.5-turbo

references to types or categories of things

oai_token-act-pair · gpt-4o-mini Triggered by @bot

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

ansas

-0.71

LESS

-0.67

Shut

-0.63

MAP

-0.63

CD

-0.61

hole

-0.60

Downloadha

-0.59

idav

-0.58

 Goodbye

-0.56

adan

-0.56

POSITIVE LOGITS

 magnitude

1.48

 caliber

1.43

 calib

1.36

 nature

1.32

 stature

1.30

 importance

1.24

 proportions

1.23

il

1.23

 size

1.20

 sorts

1.17

Activations Density 0.179%

phrases referring to certain characteristics or types mentioned in a comparison

references to types or categories of things

No Comments

No Known Activations