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

Head Attr Weights

0:0.10

1:0.07

2:0.09

3:0.07

4:0.09

5:0.09

6:0.06

7:0.08

8:0.06

9:0.07

10:0.08

11:0.10

Negative Logits

rolet

-1.25

ynes

-1.23

 upwards

-1.22

ilee

-1.21

 Rally

-1.18

 Liberty

-1.18

izont

-1.18

 obtained

-1.17

[(

-1.17

ieth

-1.16

POSITIVE LOGITS

 Shard

1.77

DERR

1.46

ktop

1.33

tics

1.30

 captcha

1.28

imei

1.28

 Byzantine

1.26

henko

1.25

qus

1.23

 WARN

1.22

Activations Density 0.000%

No Known Activations

This feature has no known activations.