INDEX

Explanations

No Explanations Found

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

Not in Any Lists

No Comments

Head Attr Weights

0:0.07

1:0.07

2:0.08

3:0.08

4:0.09

5:0.07

6:0.06

7:0.09

8:0.10

9:0.07

10:0.08

11:0.09

Negative Logits

cffffcc

-1.50

 cler

-1.49

 prost

-1.38

itudes

-1.34

aughs

-1.30

 volunteers

-1.29

atics

-1.29

 fellows

-1.27

ty

-1.26

 electrodes

-1.24

POSITIVE LOGITS

qus

1.37

MU

1.37

 warr

1.33

 Liga

1.33

 crate

1.29

meter

1.28

龍喚士

1.27

▬

1.23

ught

1.21

 Alchemist

1.20

Activations Density 0.000%

No Known Activations

This feature has no known activations.