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

phrine

-0.84

 Orig

-0.73

NetMessage

-0.67

itol

-0.66

naire

-0.65

resents

-0.65

å£«

-0.65

prints

-0.65

raviolet

-0.65

entanyl

-0.64

POSITIVE LOGITS

NRS

0.83

atory

0.82

ESA

0.78

BAT

0.70

Ø³

0.68

ain

0.67

usc

0.66

orically

0.64

NES

0.64

sat

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.