INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

UFF

-0.69

verage

-0.66

 skelet

-0.66

Dream

-0.66

acements

-0.65

alg

-0.64

Chart

-0.63

 Label

-0.63

anguages

-0.62

 Fiction

-0.61

POSITIVE LOGITS

 Blessed

0.70

ĪĴ

0.67

peria

0.65

 Shank

0.64

 Laksh

0.64

avin

0.60

ItemImage

0.59

 Walters

0.59

ounty

0.59

htaking

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.