INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

Appearances

-0.72

ALK

-0.71

iability

-0.70

gat

-0.68

escal

-0.68

loading

-0.67

telling

-0.66

itely

-0.66

lot

-0.65

iday

-0.65

POSITIVE LOGITS

ornia

1.08

 destro

0.68

ritch

0.67

 captcha

0.67

 Palo

0.66

wcs

0.65

 tiss

0.65

 completes

0.63

idav

0.62

 Stanford

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.