INDEX

Explanations

No Explanations Found

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ancial

-0.76

RIP

-0.74

sighted

-0.74

 Bucks

-0.65

 Aberdeen

-0.65

 commentary

-0.64

 marching

-0.63

 Slater

-0.63

ertodd

-0.62

 attaching

-0.61

POSITIVE LOGITS

ioxide

0.79

 cultiv

0.76

oldown

0.74

\">

0.74

enei

0.74

agin

0.73

omal

0.70

lyak

0.69

imester

0.69

roma

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.