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

ãĤ¹ãĥĪ

-0.78

çĶ°

-0.68

orate

-0.67

éĹĺ

-0.66

 Devi

-0.65

gio

-0.65

hari

-0.65

bett

-0.64

demon

-0.64

SCP

-0.63

POSITIVE LOGITS

 appre

0.65

attery

0.63

esity

0.61

otten

0.60

lamm

0.59

 Lange

0.59

hots

0.58

ofi

0.58

ignt

0.57

 sails

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.