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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 broadband

-0.73

bats

-0.69

 submarine

-0.68

 satellites

-0.66

socket

-0.65

 archived

-0.64

 hepatitis

-0.62

ADS

-0.62

 submarines

-0.62

mill

-0.62

POSITIVE LOGITS

efe

0.80

igne

0.71

ifacts

0.67

nen

0.67

ources

0.66

oses

0.65

onz

0.65

 Emer

0.64

utsu

0.64

emi

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.