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

 Sparkle

-0.75

secondary

-0.68

igne

-0.63

AE

-0.63

Ark

-0.63

 Pradesh

-0.63

 plane

-0.62

 reserves

-0.62

birds

-0.61

fecture

-0.60

POSITIVE LOGITS

glers

0.71

emonic

0.69

 endings

0.65

ategory

0.63

 penn

0.63

obiles

0.62

irtual

0.62

thur

0.62

pci

0.61

 Canaver

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.