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

Negative Logits

oros

-0.81

NAT

-0.75

RAW

-0.66

Ô

-0.65

°

-0.64

mber

-0.64

 curs

-0.63

 inflammatory

-0.62

ERT

-0.61

 Sponsor

-0.60

POSITIVE LOGITS

romeda

0.73

 Guan

0.70

illas

0.68

 Misc

0.68

hani

0.66

suits

0.64

 guiName

0.63

 Females

0.63

 Attributes

0.61

angers

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.