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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

senal

-0.76

natureconservancy

-0.69

yles

-0.68

interstitial

-0.65

 hype

-0.64

 Metroid

-0.64

 Whis

-0.63

cells

-0.63

xus

-0.62

abetes

-0.61

POSITIVE LOGITS

sburg

0.79

ritical

0.72

"]=>

0.72

.--

0.67

ï¸ı

0.66

atin

0.65

.–

0.65

onto

0.64

OUR

0.63

TO

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.