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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

asin

-0.69

 spons

-0.67

 berth

-0.66

 mortg

-0.65

contained

-0.65

olina

-0.64

"},"

-0.64

ornia

-0.63

 enrol

-0.63

lic

-0.63

POSITIVE LOGITS

SPONSORED

0.80

HR

0.68

Screen

0.68

arium

0.67

Motion

0.66

OUN

0.64

ãĥ³ãĤ¸

0.63

 Quartz

0.63

ãģĤ

0.63

 Rouge

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.