INDEX

Explanations

mentions of "rare" or "unique" items as well as significant updates related to them

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L10_Hcat_z_lr1.20e-03_l11.30e+00_ds24576_bs4096_dc1.00e-05_rsanthropic_rie25000_nr4_v9.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

float32

Hook Name

blocks.10.attn.hook_z

Hook Layer

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.07

1:0.26

2:0.04

3:0.09

4:0.09

5:0.02

6:0.10

7:0.07

8:0.03

9:0.03

10:0.06

11:0.08

Negative Logits

 Whit

-2.72

auga

-2.39

webkit

-2.34

 Directions

-2.33

xual

-2.32

 volunt

-2.32

entric

-2.31

 orientation

-2.30

deg

-2.28

 Compass

-2.18

POSITIVE LOGITS

Rare

7.24

 Rare

7.11

 rarity

6.41

 rare

6.33

Uncommon

6.09

 Uncommon

5.88

 uncommon

5.52

 unheard

4.07

Common

4.00

Legendary

4.00

Activations Density 0.013%

mentions of "rare" or "unique" items as well as significant updates related to them

No Comments

No Known Activations