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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

iqueness

-0.77

fters

-0.75

Â·Â·

-0.73

ãĤ´ãĥ³

-0.72

 skewed

-0.69

primary

-0.69

¥ŀ

-0.66

icably

-0.64

secondary

-0.63

yssey

-0.62

POSITIVE LOGITS

Pat

0.73

 Beir

0.73

 Patt

0.69

anny

0.68

DR

0.68

asia

0.66

Fel

0.65

 Dare

0.65

 Registered

0.64

 Richie

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.