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

 Colonial

-0.65

 Anthrop

-0.64

 Allen

-0.63

 Lima

-0.62

 Smithsonian

-0.61

 outpatient

-0.60

 Sunder

-0.60

 Bethlehem

-0.60

 Editorial

-0.60

igham

-0.59

POSITIVE LOGITS

MO

0.77

etheless

0.70

mo

0.67

ETF

0.66

rub

0.66

iannopoulos

0.66

gay

0.65

GR

0.63

»Ĵ

0.62

atical

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.