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

 habitable

-0.67

 Decl

-0.64

 sovere

-0.64

otyp

-0.62

 Practices

-0.62

 Colon

-0.60

Ada

-0.60

 expans

-0.59

asive

-0.59

 Discrimination

-0.59

POSITIVE LOGITS

Haunted

0.77

fleet

0.75

ssh

0.73

AAAAAAAA

0.72

gow

0.72

itary

0.72

BB

0.72

RW

0.69

uve

0.68

rar

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.