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

ULAR

-0.75

Shut

-0.71

shut

-0.70

ATA

-0.68

uce

-0.68

 travel

-0.68

 spiral

-0.67

ET

-0.66

 habit

-0.64

 shuttle

-0.63

POSITIVE LOGITS

akespe

0.77

ormons

0.75

 ILCS

0.75

psc

0.74

aughs

0.70

 ..............

0.65

iban

0.65

 Okawaru

0.65

icians

0.65

ews

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.