INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

addons

-0.71

 fetch

-0.70

 retract

-0.69

TNT

-0.68

 underground

-0.67

 setting

-0.65

tics

-0.65

ä»£

-0.65

 channels

-0.64

cart

-0.62

POSITIVE LOGITS

arten

0.85

erg

0.81

reciation

0.76

osity

0.75

ALTH

0.72

bernatorial

0.72

Kyl

0.71

unn

0.71

ENGTH

0.71

assi

0.70

Activations Density 0.000%

No Known Activations

This feature has no known activations.