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

rika

-0.75

 Myster

-0.71

 Magazine

-0.65

natureconservancy

-0.64

Ref

-0.64

imaru

-0.64

\":

-0.62

apest

-0.61

 Instr

-0.59

 Thousand

-0.58

POSITIVE LOGITS

ocy

0.71

icum

0.71

efe

0.69

cade

0.68

spection

0.68

akings

0.67

worker

0.66

 Cheong

0.64

essions

0.63

arthed

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.