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

quickShipAvailable

-0.72

iband

-0.70

 Elect

-0.67

 credential

-0.64

insula

-0.64

 deleg

-0.63

Deal

-0.61

 Booster

-0.60

ammy

-0.60

 Debor

-0.59

POSITIVE LOGITS

\/

0.75

ces

0.70

bane

0.69

psc

0.69

ONES

0.69

doms

0.68

 Stras

0.66

monds

0.66

Arsenal

0.65

wordpress

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.