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

 fold

-0.71

rave

-0.65

axis

-0.64

isable

-0.63

terday

-0.63

avez

-0.62

 Subway

-0.60

 truth

-0.59

erial

-0.59

 aisle

-0.58

POSITIVE LOGITS

 Priebus

0.70

annis

0.67

NetMessage

0.64

ahime

0.63

 Merrill

0.62

artifacts

0.62

iciency

0.61

 Cunning

0.61

ples

0.60

 Scully

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.