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

 Dayton

-0.77

ppard

-0.74

aceous

-0.69

hement

-0.66

Rated

-0.64

ommod

-0.63

 Brooks

-0.59

 Rochester

-0.59

igham

-0.59

 Wichita

-0.58

POSITIVE LOGITS

 simultaneously

1.36

0.90

 concurrently

0.88

 Awakens

0.78

Ñı

0.78

 ãĤ

0.77

ì

0.71

terday

0.70

 simultane

0.70

acters

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.