INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

anship

-0.73

ffe

-0.67

gged

-0.60

mble

-0.60

ropolis

-0.58

OPLE

-0.58

iji

-0.58

ucket

-0.58

ouver

-0.57

lda

-0.56

POSITIVE LOGITS

2.74

sb

1.30

sburg

1.27

ski

1.27

sat

1.24

sin

1.23

ses

1.22

sis

1.21

sa

1.21

sf

1.13

Activations Density 0.000%

No Known Activations

This feature has no known activations.