INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

jriwal

-0.75

zeb

-0.74

 reckoning

-0.69

 Sunder

-0.69

 footing

-0.67

 sleeping

-0.67

 flush

-0.66

 stra

-0.64

ple

-0.62

 spons

-0.61

POSITIVE LOGITS

recated

0.80

oise

0.71

ĳå£«

0.69

ĺ

0.68

hers

0.68

TED

0.67

 Opera

0.67

aspers

0.67

cap

0.67

manship

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.