INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 cryst

-0.79

mercial

-0.75

 experimenting

-0.68

ailability

-0.66

 advent

-0.65

 harvesting

-0.64

 nesting

-0.63

 manipulating

-0.63

 arrang

-0.62

 pacing

-0.62

POSITIVE LOGITS

quickShipAvailable

0.74

izens

0.72

 Wheels

0.70

 Fault

0.67

achu

0.67

erick

0.66

ãĤ¼

0.66

eez

0.66

OAD

0.64

argon

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.