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

obos

-0.71

=-=-=-=-

-0.70

pload

-0.69

eph

-0.68

ldom

-0.67

units

-0.66

shared

-0.66

apore

-0.66

eus

-0.65

avia

-0.63

POSITIVE LOGITS

buster

0.66

 Nationwide

0.62

 duly

0.61

 confisc

0.60

busters

0.58

ently

0.58

 religiously

0.57

CSI

0.57

 Mori

0.56

speak

0.56

Activations Density 0.000%

No Known Activations

This feature has no known activations.