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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ulp

-0.68

orem

-0.67

UCT

-0.66

 Mant

-0.65

 Gibson

-0.65

ItemTracker

-0.64

OWS

-0.64

berries

-0.63

 Cantor

-0.63

****************

-0.62

POSITIVE LOGITS

olation

0.81

 Affairs

0.78

oÄŁ

0.68

 aggrav

0.66

 transitioned

0.63

 transitioning

0.62

cele

0.62

ney

0.61

FACE

0.60

ĺħ

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.