INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

nels

-0.67

NK

-0.66

 Builder

-0.66

 redes

-0.65

atorium

-0.65

 Observer

-0.64

 Raphael

-0.64

 Jonah

-0.63

 Gardner

-0.63

 Webb

-0.62

POSITIVE LOGITS

renheit

0.77

grain

0.76

gments

0.76

è¦ļéĨĴ

0.74

trak

0.74

ntil

0.73

arine

0.73

ignty

0.70

vous

0.70

fortune

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.