INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

yond

-0.84

arov

-0.73

itored

-0.72

voy

-0.72

TPPStreamerBot

-0.71

spirit

-0.69

zh

-0.68

ording

-0.67

eren

-0.66

monary

-0.66

POSITIVE LOGITS

 destro

0.76

 promoter

0.68

FX

0.63

MFT

0.62

DI

0.62

recated

0.61

 Rena

0.61

AAA

0.61

FAT

0.60

EED

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.