INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

ront

-0.83

akis

-0.66

bian

-0.61

bh

-0.60

otiation

-0.60

iferation

-0.60

 anticipation

-0.59

aspberry

-0.59

ocolate

-0.58

inav

-0.57

POSITIVE LOGITS

©¶æ¥µ

0.77

CoC

0.74

 Clicker

0.72

urga

0.72

£ı

0.72

 Zeit

0.72

 Turks

0.69

 Skydragon

0.68

Kun

0.67

KDE

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.