INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

oji

-0.82

imeters

-0.70

 Pearce

-0.66

 grading

-0.66

 ceilings

-0.63

 indefinite

-0.62

 captcha

-0.62

 Anniversary

-0.61

 commuting

-0.61

esters

-0.61

POSITIVE LOGITS

BIL

0.84

Bet

0.72

Bush

0.70

ADA

0.67

CH

0.65

DragonMagazine

0.65

GH

0.64

ENE

0.64

dl

0.64

Fred

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.