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

alty

-0.78

eger

-0.74

 Triangle

-0.68

haar

-0.65

ttes

-0.65

²¾

-0.63

 peers

-0.61

Element

-0.61

rences

-0.59

ilion

-0.59

POSITIVE LOGITS

 Skydragon

0.71

TERN

0.68

ovember

0.65

odd

0.65

adium

0.62

ovo

0.62

KP

0.59

UX

0.59

ngth

0.59

 postpone

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.