INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ulative

-0.78

ngth

-0.74

hari

-0.73

VIDIA

-0.70

Works

-0.69

ITH

-0.69

ĸļ

-0.68

ividual

-0.68

veyard

-0.67

hs

-0.67

POSITIVE LOGITS

 elusive

0.67

RFC

0.67

 Myanmar

0.67

Topic

0.65

 mmol

0.64

 precursor

0.64

ÙĲ

0.64

oid

0.64

pur

0.62

alogue

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.