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

Head Attr Weights

0:0.07

1:0.08

2:0.09

3:0.07

4:0.10

5:0.07

6:0.08

7:0.07

8:0.08

9:0.07

10:0.10

11:0.07

Negative Logits

 equivalents

-1.37

atten

-1.34

-1.28

 duplication

-1.28

 consultancy

-1.25

omet

-1.23

��

-1.22

 Duff

-1.21

 patented

-1.20

POSITIVE LOGITS

andals

1.56

etc

1.49

 IMAGES

1.36

ⓘ

1.36

 UFOs

1.34

 MISS

1.32

ithing

1.30

ⓘ

1.28

 Strikes

1.27

comings

1.26

Activations Density 0.000%

No Known Activations

This feature has no known activations.