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

ACA

-0.70

err

-0.65

etsk

-0.63

çĶ

-0.63

awaited

-0.62

inas

-0.58

Fourth

-0.58

pring

-0.58

 Crescent

-0.58

 forecast

-0.58

POSITIVE LOGITS

adena

0.78

BuyableInstoreAndOnline

0.77

argon

0.75

VICE

0.74

ATURE

0.71

oxide

0.70

 Scient

0.70

kamp

0.68

 Crusher

0.68

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.