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

udo

-0.86

NRS

-0.81

ateurs

-0.76

"]=>

-0.75

 shortages

-0.71

aida

-0.70

osate

-0.67

ague

-0.67

Lv

-0.66

ambo

-0.66

POSITIVE LOGITS

Ĥİ

0.89

inventoryQuantity

0.65

 abort

0.64

İĭ

0.63

 guiActiveUnfocused

0.61

ãĥĥãĥī

0.60

 Chatt

0.58

forth

0.56

wake

0.56

hower

0.56

Activations Density 0.000%

No Known Activations

This feature has no known activations.