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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Nich

-0.73

 Tiff

-0.72

©¶æ

-0.68

 Drew

-0.67

 Prelude

-0.66

Tar

-0.64

 Nationwide

-0.64

 Matt

-0.62

 casc

-0.61

POSITIVE LOGITS

 intelligence

1.39

 Intelligence

1.27

elligence

1.06

intelligence

1.03

dinand

0.81

iable

0.79

 dexterity

0.78

elligent

0.77

urion

0.77

ieth

0.76

Activations Density 0.000%

No Known Activations

This feature has no known activations.