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

Ny

-0.76

 Bottom

-0.64

NL

-0.61

 Nether

-0.59

Nep

-0.59

quote

-0.58

 McKenzie

-0.56

 Martian

-0.56

 Sahara

-0.56

NS

-0.56

POSITIVE LOGITS

ween

0.79

Magikarp

0.71

vernment

0.71

itect

0.71

States

0.69

ebted

0.68

ERAL

0.68

byss

0.68

uations

0.67

apore

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.