INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L1_Hcat_z_lr1.20e-03_l18.00e-01_ds24576_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v5.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

float32

Hook Name

blocks.1.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

0:0.08

1:0.08

2:0.08

3:0.08

4:0.08

5:0.09

6:0.08

7:0.08

8:0.07

9:0.07

10:0.07

11:0.07

Negative Logits

CLASSIFIED

-1.79

thood

-1.76

akespeare

-1.51

ql

-1.41

eworks

-1.33

iscons

-1.32

 Verse

-1.30

requisite

-1.27

college

-1.26

=]

-1.25

POSITIVE LOGITS

 Rodrig

1.31

 Kron

1.29

 Greenwood

1.28

 Labrador

1.27

 Canaver

1.25

 Bark

1.25

 Dangerous

1.24

rou

1.24

 Baal

1.22

 venom

1.21

Activations Density 0.000%

No Known Activations

This feature has no known activations.