INDEX

Explanations

titles, names, and honorifics associated with dignitaries or individuals in positions of authority

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L5_Hcat_z_lr1.20e-03_l11.00e+00_ds49152_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

float32

Hook Name

blocks.5.attn.hook_z

Hook Layer

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

IFrame

Link

Head Attr Weights

0:0.02

1:0.02

2:0.07

3:0.04

4:0.05

5:0.03

6:0.53

7:0.04

8:0.02

9:0.04

10:0.06

11:0.03

Negative Logits

inea

-1.48

Reviewer

-1.45

��

-1.42

acea

-1.39

 steroid

-1.35

イ

-1.29

ा

-1.27

STD

-1.25

�

-1.25

 piping

-1.23

POSITIVE LOGITS

���

1.58

ilts

1.43

paren

1.38

former

1.35

aughter

1.31

 Lann

1.31

YP

1.30

Ott

1.29

asury

1.28

ongh

1.27

Activations Density 0.029%

titles, names, and honorifics associated with dignitaries or individuals in positions of authority

No Known Activations