INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L5_Hcat_z_lr1.20e-03_l11.00e+00_ds49152_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

float32

Hook Name

blocks.5.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

1:0.06

2:0.08

3:0.09

4:0.09

5:0.08

6:0.07

7:0.07

8:0.07

9:0.09

10:0.08

11:0.08

Negative Logits

 Morty

-1.65

DragonMagazine

-1.56

 Titans

-1.56

 Turtles

-1.48

 =================================

-1.48

Rail

-1.46

RED

-1.46

ESL

-1.44

ELE

-1.43

 nerds

-1.43

POSITIVE LOGITS

 intervening

1.58

oward

1.53

etting

1.52

cho

1.50

biased

1.45

ban

1.39

recent

1.38

married

1.38

 Documents

1.37

avorable

1.36

Activations Density 0.000%

No Known Activations

This feature has no known activations.