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

 onward

-0.70

 distribut

-0.66

 Torrent

-0.66

izont

-0.65

onym

-0.64

Doc

-0.63

roller

-0.62

 partName

-0.62

MO

-0.61

loo

-0.60

POSITIVE LOGITS

ise

1.04

 unwanted

0.77

ISE

0.76

wav

0.70

thritis

0.69

bestos

0.67

 seiz

0.66

ktop

0.65

ersive

0.64

strument

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.