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

bryce

-0.71

 WATCHED

-0.71

 Instructor

-0.68

aeda

-0.68

 Assist

-0.68

 Lanka

-0.68

 Equip

-0.66

":[

-0.66

zsche

-0.65

men

-0.64

POSITIVE LOGITS

ibles

0.70

ible

0.65

storage

0.62

 clouds

0.61

assium

0.60

axy

0.59

illard

0.59

DO

0.58

signed

0.58

 grapes

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.