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

plate

-0.73

enger

-0.71

owler

-0.70

 Span

-0.63

 Lash

-0.63

 passer

-0.62

olk

-0.61

bird

-0.61

bomb

-0.61

âĢİ

-0.60

POSITIVE LOGITS

ahime

0.77

uana

0.77

soDeliveryDate

0.69

ItemTracker

0.69

isSpecialOrderable

0.68

NK

0.67

accompan

0.65

netflix

0.64

©¶æ¥µ

0.64

Ire

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.