INDEX

Explanations

phrases related to approval or authorization

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

0:0.06

1:0.02

2:0.20

3:0.06

4:0.28

5:0.05

6:0.02

7:0.02

8:0.07

9:0.11

10:0.04

11:0.01

Negative Logits

ussen

-1.54

USS

-1.46

��

-1.44

 Cous

-1.40

��

-1.40

�

-1.39

�

-1.38

���

-1.35

��極

-1.31

Vaj

-1.30

POSITIVE LOGITS

ioxide

1.72

oxide

1.60

 gases

1.31

 emitting

1.31

 aspir

1.30

gre

1.29

 oxide

1.27

oak

1.27

IDE

1.25

 forests

1.24

Activations Density 0.008%

phrases related to approval or authorization

No Known Activations