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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

FactoryReloaded

-0.83

âĸ¬

-0.75

fecture

-0.67

Reloaded

-0.63

Radio

-0.62

 Survival

-0.62

arthed

-0.62

 Masquerade

-0.62

mite

-0.62

 Burning

-0.61

POSITIVE LOGITS

essage

0.74

alian

0.71

umsy

0.68

 prosec

0.68

ÅĤ

0.68

iston

0.68

iano

0.67

awan

0.66

iscons

0.66

'[

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.