INDEX

Explanations

No Explanations Found

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

Not in Any Lists

No Comments

Head Attr Weights

0:0.07

1:0.08

2:0.07

3:0.09

4:0.08

5:0.07

6:0.09

7:0.07

8:0.08

9:0.07

10:0.10

11:0.07

Negative Logits

 recept

-1.51

 Orient

-1.34

ynt

-1.32

Bottom

-1.29

ヘラ

-1.28

 Citation

-1.28

 Contents

-1.25

enses

-1.24

 widgets

-1.24

 Adapter

-1.20

POSITIVE LOGITS

RIP

1.33

abama

1.32

omics

1.31

HQ

1.31

agnar

1.28

iversary

1.26

Dispatch

1.21

!),

1.20

icum

1.20

zyk

1.19

Activations Density 0.000%

No Known Activations

This feature has no known activations.