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

kj

-0.80

*/(

-0.78

ikarp

-0.77

bek

-0.69

Ã§a

-0.68

¶ħ

-0.67

Â¶

-0.67

laden

-0.67

beck

-0.66

felt

-0.66

POSITIVE LOGITS

unction

0.79

 Scorpion

0.68

 CASE

0.67

 Tamil

0.66

oons

0.65

 Minority

0.63

 Spac

0.61

 priests

0.60

emetery

0.60

visor

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.