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

ady

-0.68

bring

-0.67

iversal

-0.67

ascus

-0.63

Deal

-0.62

achus

-0.60

cles

-0.60

speak

-0.60

 elbows

-0.58

 pans

-0.57

POSITIVE LOGITS

£ı

0.88

ãĥ´

0.80

ku

0.71

 periodic

0.70

å£«

0.69

lement

0.68

ensation

0.66

KI

0.66

ination

0.65

VPN

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.