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

artifacts

-0.74

stretched

-0.73

ãĥĺ

-0.71

mast

-0.68

moil

-0.67

Zub

-0.66

ka

-0.64

beh

-0.63

Kaf

-0.62

stood

-0.60

POSITIVE LOGITS

essee

0.74

Marie

0.71

ovember

0.69

enth

0.69

orney

0.66

iatus

0.65

 Description

0.64

isse

0.62

ivia

0.62

lectic

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.