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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ortium

-0.69

lance

-0.68

 Andrews

-0.67

azaki

-0.67

emale

-0.67

Elf

-0.67

Ryan

-0.67

za

-0.63

 Converted

-0.63

¬¼

-0.63

POSITIVE LOGITS

fr

0.85

phys

0.77

resp

0.68

 psychiat

0.64

gy

0.64

prov

0.63

 fingerprints

0.63

 inhibitors

0.63

hid

0.63

hog

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.