INDEX

Explanations

references to church or religious institutions and their activities

oai_token-act-pair · gpt-4o-mini Triggered by @bot

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

0:0.28

1:0.01

2:0.17

3:0.08

4:0.03

5:0.06

6:0.03

7:0.06

8:0.16

9:0.02

10:0.03

11:0.02

Negative Logits

 Droid

-3.25

 GoPro

-3.23

 Xiaomi

-3.23

 Gadget

-3.16

 Cyborg

-3.15

UID

-3.01

 droid

-2.85

uez

-2.80

 Verge

-2.78

 Garmin

-2.78

POSITIVE LOGITS

 congregation

6.26

 pastor

5.45

 church

5.44

 pastors

5.37

 congreg

5.28

 parish

5.21

 churches

5.21

 sermon

5.12

 Episcopal

5.04

church

5.02

Activations Density 0.149%

references to church or religious institutions and their activities

No Comments

No Known Activations