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

ielding

-0.74

pour

-0.71

clair

-0.69

anting

-0.66

ntil

-0.66

elong

-0.66

ifully

-0.64

pen

-0.64

fman

-0.63

anca

-0.62

POSITIVE LOGITS

 Monstrous

0.74

URR

0.74

RU

0.72

rio

0.70

hr

0.66

\\\\\\\\\\\\\\\\

0.66

inventoryQuantity

0.64

ptroller

0.63

 ages

0.63

 CLSID

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.