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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Freud

-0.72

 Huntington

-0.68

chem

-0.63

aban

-0.63

aman

-0.63

DW

-0.62

 Blacks

-0.62

EStreamFrame

-0.62

roph

-0.59

alysis

-0.59

POSITIVE LOGITS

uge

0.72

INST

0.64

ovi

0.62

WAY

0.62

ppelin

0.61

received

0.61

UGE

0.60

().

0.59

shape

0.59

olute

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.