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

 Criminal

-0.70

CPS

-0.69

 faculty

-0.69

ONSORED

-0.67

 Investigative

-0.66

Ark

-0.64

 exoner

-0.62

FactoryReloaded

-0.62

 Argon

-0.61

 wrongful

-0.61

POSITIVE LOGITS

atz

0.75

ascript

0.74

zsche

0.73

ilon

0.72

tests

0.71

liga

0.70

igers

0.69

acha

0.69

staking

0.67

intendent

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.