INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Verse

-0.75

Border

-0.74

aeus

-0.72

ãĥŃ

-0.70

ãĥĥ

-0.70

ma

-0.69

iel

-0.68

 Tome

-0.67

ãĤ¢

-0.67

 veget

-0.67

POSITIVE LOGITS

 Wasserman

0.78

 probes

0.71

 examinations

0.70

JPM

0.69

chers

0.69

ISI

0.68

pse

0.68

 elect

0.67

 corro

0.66

 Palest

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.