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

PIN

-0.84

Redd

-0.80

plot

-0.80

credit

-0.76

Merit

-0.73

Uber

-0.71

Thumbnail

-0.69

Magikarp

-0.69

Apps

-0.69

hedon

-0.68

POSITIVE LOGITS

 Accord

0.63

 Ernst

0.63

ilater

0.62

 certified

0.62

stand

0.62

lap

0.62

atile

0.61

swing

0.60

ilion

0.60

heed

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.