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

romeda

-0.83

apolis

-0.79

emet

-0.74

unal

-0.72

UTE

-0.70

ori

-0.70

jp

-0.67

 Airl

-0.66

ocratic

-0.65

ASE

-0.65

POSITIVE LOGITS

©¶æ

0.79

hatt

0.73

wolf

0.73

horn

0.71

 Sass

0.70

restricted

0.66

 borrower

0.66

currency

0.66

speed

0.65

shots

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.