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

oris

-0.68

webkit

-0.65

 Giovanni

-0.63

mber

-0.63

zu

-0.63

cture

-0.62

ggle

-0.61

 Lion

-0.60

Ott

-0.60

WP

-0.58

POSITIVE LOGITS

senal

0.82

pload

0.78

ividual

0.76

selves

0.76

subject

0.69

ructose

0.68

alties

0.68

âĵĺ

0.67

itent

0.67

DragonMagazine

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.