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

 Bernstein

-0.73

 fitted

-0.66

abase

-0.66

rav

-0.66

 conv

-0.63

 booked

-0.63

 segregated

-0.63

Das

-0.62

vit

-0.61

 Lindsey

-0.61

POSITIVE LOGITS

BuyableInstoreAndOnline

0.72

CONT

0.70

RON

0.70

*/(

0.67

addons

0.66

 annoyance

0.65

metadata

0.65

ãĥĩãĤ£

0.65

 Leilan

0.65

,,,,

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.