INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Emin

-0.70

dom

-0.70

âĹ¼

-0.65

doms

-0.65

nep

-0.62

boss

-0.62

ean

-0.61

obl

-0.59

Ń·

-0.59

Eng

-0.58

POSITIVE LOGITS

ption

0.72

 glim

0.68

isphere

0.68

enough

0.68

 Pratt

0.65

ana

0.64

achine

0.63

lia

0.63

ably

0.62

llo

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.