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

DS

-0.67

BRA

-0.66

Ble

-0.65

USA

-0.65

SEA

-0.63

Ga

-0.63

Desk

-0.63

hr

-0.61

da

-0.61

 Tribune

-0.60

POSITIVE LOGITS

imental

0.83

anmar

0.69

lehem

0.68

iosyncr

0.67

 eleph

0.67

iless

0.67

ilitary

0.66

rices

0.66

uyomi

0.66

ruary

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.