INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

edes

-0.74

 Finch

-0.70

angular

-0.68

aea

-0.66

isphere

-0.64

foreign

-0.64

 Essex

-0.63

fetched

-0.63

cia

-0.61

agent

-0.59

POSITIVE LOGITS

 Reloaded

0.68

ISTER

0.64

HIP

0.63

Ø§

0.63

è»

0.62

©¶æ

0.62

 Bunny

0.61

à©

0.60

SQU

0.59

æ³

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.