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

xual

-0.78

 convol

-0.72

 sqor

-0.70

 collectively

-0.70

RAL

-0.69

SpaceEngineers

-0.68

draft

-0.68

STAT

-0.68

 resid

-0.68

ACTED

-0.65

POSITIVE LOGITS

roo

0.70

ments

0.68

iche

0.68

wer

0.67

robe

0.66

ando

0.65

hen

0.63

zees

0.62

leon

0.62

sson

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.