INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L9_Hcat_z_lr1.20e-03_l11.20e+00_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.9.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

1:0.09

2:0.07

3:0.06

4:0.08

5:0.08

6:0.07

7:0.07

8:0.10

9:0.08

10:0.09

11:0.08

Negative Logits

 cord

-2.88

 Bleach

-2.59

ted

-2.49

vac

-2.46

ovies

-2.45

strate

-2.42

sep

-2.40

CL

-2.38

blank

-2.28

��

-2.28

POSITIVE LOGITS

edin

3.10

 Erit

2.66

 Motorsport

2.58

ECB

2.56

 CentOS

2.55

Ö

2.54

Els

2.48

 Macedonia

2.39

 Bundes

2.34

Erd

2.33

Activations Density 0.000%

No Known Activations

This feature has no known activations.