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

iction

-0.70

Ö

-0.63

 lymph

-0.62

dyl

-0.62

 Cadillac

-0.60

 Chevrolet

-0.59

rontal

-0.58

bill

-0.58

ãĥ´

-0.58

ictions

-0.58

POSITIVE LOGITS

anical

0.69

 Breaker

0.69

essen

0.68

Export

0.67

Ao

0.67

PLA

0.66

VALUE

0.66

anwhile

0.64

 Rowling

0.63

ORPG

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.