INDEX

Explanations

shapes and forms within design contexts

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L2_Hcat_z_lr1.20e-03_l11.00e+00_ds24576_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v4.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

float32

Hook Name

blocks.2.attn.hook_z

Hook Layer

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

IFrame

Link

Head Attr Weights

0:0.05

1:0.04

2:0.16

3:0.06

4:0.04

5:0.08

6:0.03

7:0.03

8:0.11

9:0.26

10:0.06

11:0.03

Negative Logits

heimer

-1.46

achev

-1.43

reb

-1.42

heim

-1.41

)</

-1.39

owitz

-1.36

),"

-1.35

lease

-1.35

)",

-1.35

mun

-1.34

POSITIVE LOGITS

 scrolling

1.39

fog

1.33

tra

1.24

 scan

1.18

 Blend

1.17

 gradient

1.16

 veil

1.16

 science

1.14

 Tropical

1.14

 breeze

1.13

Activations Density 0.011%

shapes and forms within design contexts

No Known Activations