INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

iffe

-0.78

demand

-0.78

eem

-0.74

FOX

-0.69

xus

-0.69

emo

-0.69

ç¥ŀ

-0.66

liga

-0.65

nee

-0.64

lar

-0.64

POSITIVE LOGITS

Distance

0.73

#$

0.70

 itch

0.70

 correspondence

0.69

////////////////

0.64

======

0.64

ãĤ¦ãĤ¹

0.63

////////

0.62

NetMessage

0.62

akings

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.