INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

erous

-0.89

eatures

-0.82

ategor

-0.78

itory

-0.70

ums

-0.70

":["

-0.68

haps

-0.67

ctic

-0.67

umerable

-0.66

ourse

-0.66

POSITIVE LOGITS

port

0.99

 Prison

0.87

Luck

0.70

 refuse

0.67

 harass

0.65

Rush

0.65

 motivations

0.65

Draft

0.64

 assassinated

0.62

 Barkley

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.