INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 successes

-0.70

paren

-0.68

podcast

-0.66

cedented

-0.65

 nightmares

-0.64

BBC

-0.64

earable

-0.63

harm

-0.63

 behold

-0.63

ÃĽ

-0.62

POSITIVE LOGITS

iage

0.83

 Doodle

0.80

 Cheong

0.78

 Eaton

0.75

 Kear

0.71

ivan

0.70

 Conver

0.69

 Dinosaur

0.68

 Chim

0.68

ioxide

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.