INDEX

Explanations

mentions of wars and military conflicts

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L10_Hcat_z_lr1.20e-03_l11.30e+00_ds24576_bs4096_dc1.00e-05_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.10.attn.hook_z

Hook Layer

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

IFrame

Link

Head Attr Weights

0:0.11

1:0.02

2:0.03

3:0.08

4:0.05

5:0.02

6:0.08

7:0.34

8:0.01

9:0.05

10:0.03

11:0.12

Negative Logits

club

-2.01

volt

-1.99

 gang

-1.98

 Warlock

-1.90

Boh

-1.89

Beast

-1.89

Sport

-1.88

WWE

-1.88

 Jury

-1.86

bay

-1.83

POSITIVE LOGITS

 Dempsey

2.31

 sqor

2.19

Yen

2.03

nav

2.02

pty

1.96

 nostalg

1.95

oppy

1.91

パ

1.91

 Leban

1.91

fff

1.89

Activations Density 0.125%

mentions of wars and military conflicts

No Known Activations