INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

gans

-0.69

ãĤ¶

-0.68

 Grounds

-0.68

ÃĥÃĤ

-0.65

 Pyro

-0.63

ãĥķãĤ©

-0.63

 Loaded

-0.63

 Dwarf

-0.62

 Sven

-0.61

 innocence

-0.61

POSITIVE LOGITS

rium

0.69

chapter

0.69

hur

0.68

OVA

0.67

Advertisement

0.65

 reconc

0.65

Middle

0.64

Spoiler

0.64

live

0.64

escape

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.