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

jury

-0.75

overed

-0.71

 ambush

-0.66

rome

-0.66

 turtles

-0.64

 advisory

-0.63

ods

-0.62

 acquisitions

-0.61

cre

-0.60

 dominated

-0.60

POSITIVE LOGITS

 Shap

0.83

 Alternatively

0.72

 Moreover

0.71

CHAT

0.69

quickShipAvailable

0.69

imentary

0.68

 Salon

0.67

Furthermore

0.67

ibrary

0.67

aeda

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.