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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

soDeliveryDate

-0.79

inous

-0.72

_{

-0.69

":["

-0.67

}}}

-0.66

inventoryQuantity

-0.66

NetMessage

-0.66

inki

-0.66

 Archangel

-0.65

ombies

-0.64

POSITIVE LOGITS

itial

0.69

agus

0.64

Jiu

0.63

imo

0.63

ascript

0.62

 handmade

0.62

ridor

0.62

ESA

0.62

utral

0.62

lam

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.