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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

stocks

-0.68

aunch

-0.67

ierre

-0.67

Decl

-0.66

hers

-0.66

lund

-0.65

arius

-0.62

swing

-0.62

leeve

-0.62

anche

-0.61

POSITIVE LOGITS

soDeliveryDate

0.88

BuyableInstoreAndOnline

0.74

uld

0.65

NCT

0.64

 Samantha

0.61

soType

0.60

 Medline

0.60

SEM

0.60

Gly

0.58

 Trials

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.