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

 destro

-0.94

rugged

-0.87

eele

-0.87

orsche

-0.68

iscovery

-0.66

 coma

-0.65

anas

-0.63

)",

-0.63

ored

-0.63

 practition

-0.62

POSITIVE LOGITS

liest

0.79

wives

0.73

ments

0.66

llan

0.66

lings

0.65

 looms

0.63

ment

0.63

berry

0.63

 Winds

0.63

tails

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.