INDEX

Explanations

No Explanations Found

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

 Brist

-0.70

 Bold

-0.69

 Bridge

-0.66

 Split

-0.64

 Tampa

-0.63

 behind

-0.63

BRE

-0.63

 bruising

-0.63

 Adin

-0.62

 Shore

-0.61

POSITIVE LOGITS

ICAN

0.82

articles

0.78

meier

0.78

phabet

0.75

umbledore

0.74

nesday

0.72

gins

0.72

ilitarian

0.71

emis

0.70

thumbnails

0.70

Activations Density 0.000%

No Known Activations

This feature has no known activations.