INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Mahjong

-0.76

 Compos

-0.67

imensional

-0.64

sted

-0.64

 Alone

-0.63

 Grimm

-0.62

 Jelly

-0.61

 Shape

-0.61

SCP

-0.61

 halluc

-0.60

POSITIVE LOGITS

xit

0.88

iries

0.79

mingham

0.79

isitions

0.76

uters

0.75

argon

0.73

business

0.70

baugh

0.68

regor

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.