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

 awaken

-0.74

angel

-0.73

geist

-0.72

 rall

-0.67

 respons

-0.67

 lawn

-0.67

zie

-0.66

 tire

-0.66

lod

-0.64

 tires

-0.63

POSITIVE LOGITS

CHAT

0.76

aghan

0.75

 largeDownload

0.71

"]=>

0.70

UTF

0.70

 Piercing

0.69

raltar

0.68

USC

0.67

][/

0.67

%]

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.