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

lace

-0.86

ringe

-0.76

cheat

-0.73

rition

-0.71

alg

-0.70

Ļ½

-0.70

dry

-0.68

ring

-0.68

Tokens

-0.67

roe

-0.67

POSITIVE LOGITS

HQ

0.66

 orders

0.65

SK

0.64

 Kens

0.64

 Samoa

0.63

 IMAGES

0.62

 manag

0.62

 Azerb

0.61

Nam

0.61

istan

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.