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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

orer

-0.73

 Sina

-0.68

razil

-0.67

asin

-0.64

endor

-0.62

 Irving

-0.62

amen

-0.62

rev

-0.62

oration

-0.60

 Ames

-0.60

POSITIVE LOGITS

 hemor

0.72

veyard

0.69

 attackers

0.67

istant

0.65

ãĥĻ

0.65

 anecd

0.65

Shield

0.64

 antagonist

0.64

 blockers

0.64

 distingu

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.