INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Helm

-0.72

TPPStreamerBot

-0.70

 Archdemon

-0.70

ques

-0.68

 Cree

-0.67

knit

-0.67

NEY

-0.65

lee

-0.63

 cellul

-0.63

agonal

-0.62

POSITIVE LOGITS

tymology

0.75

anon

0.68

 deduction

0.66

Ĥª

0.66

BRE

0.64

 invention

0.64

Pub

0.62

 ACTION

0.62

gemony

0.60

 Hash

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.