INDEX

Explanations

terms related to unmanned aerial vehicles and their applications

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L3_Hcat_z_lr1.20e-03_l19.00e-01_ds24576_bs4096_dc1.00e-06_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.3.attn.hook_z

Hook Layer

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

IFrame

Link

Head Attr Weights

0:0.04

1:0.03

2:0.08

3:0.13

4:0.02

5:0.03

6:0.13

7:0.18

8:0.05

9:0.03

10:0.07

11:0.15

Negative Logits

 Liqu

-1.45

 Ruin

-1.39

 Bullets

-1.36

 Corruption

-1.28

 Younger

-1.18

 Flav

-1.16

 Sultan

-1.11

 Phar

-1.06

 Plague

-1.06

utra

-1.05

POSITIVE LOGITS

ighth

1.24

operated

1.18

awei

1.14

 autonomous

1.14

�

1.12

Deploy

1.10

ailability

1.08

Got

1.06

uclear

1.05

obiles

1.04

Activations Density 0.001%

terms related to unmanned aerial vehicles and their applications

No Known Activations