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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ipers

-0.75

ovember

-0.75

igun

-0.75

icons

-0.73

rats

-0.73

ovie

-0.67

atech

-0.67

 Founders

-0.66

rypt

-0.65

ateurs

-0.65

POSITIVE LOGITS

 proble

0.69

same

0.68

WP

0.64

 tenure

0.62

married

0.61

AME

0.61

 circulate

0.61

stress

0.60

 Seym

0.60

isSpecialOrderable

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.