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

WATCH

-0.69

Dispatch

-0.67

fits

-0.67

ãĥĳ

-0.65

 GOODMAN

-0.65

icter

-0.65

Sport

-0.65

select

-0.63

 partName

-0.60

Reader

-0.60

POSITIVE LOGITS

uana

0.73

 Forbidden

0.71

 Flask

0.70

 Galactic

0.67

 Andromeda

0.63

eto

0.62

lia

0.62

 Ancients

0.62

etus

0.61

 acknowled

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.