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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hub

-0.70

VIEW

-0.69

NOR

-0.66

 validation

-0.65

opian

-0.64

 metab

-0.62

met

-0.62

 familiar

-0.62

 upside

-0.61

 bullish

-0.60

POSITIVE LOGITS

culosis

0.82

 Flavoring

0.78

 Railroad

0.71

arations

0.69

 Pension

0.69

 traged

0.69

ovie

0.68

 Railway

0.65

 ages

0.63

aughtered

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.