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

Frameworks

-0.69

 managers

-0.68

errors

-0.65

ERA

-0.64

 sovere

-0.63

undo

-0.62

 ambition

-0.62

isSpecialOrderable

-0.62

 managerial

-0.60

ndra

-0.60

POSITIVE LOGITS

meet

0.74

fed

0.71

 Posted

0.67

abis

0.66

atown

0.66

 Spread

0.66

pei

0.65

ffff

0.65

ded

0.64

 ..........

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.