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

phans

-0.82

Bid

-0.72

 Converted

-0.71

 Chaser

-0.70

 Dmit

-0.69

nis

-0.68

 Amid

-0.68

Merit

-0.66

pci

-0.65

Hai

-0.64

POSITIVE LOGITS

LINE

0.78

 predec

0.69

 MUCH

0.68

 adolesc

0.68

onder

0.67

 irresistible

0.66

 exha

0.66

 expansions

0.64

 critically

0.63

FIN

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.