INDEX

Explanations

terms related to safeguarding and security

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_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

Not in Any Lists

No Comments

Head Attr Weights

0:0.14

1:0.03

2:0.10

3:0.03

4:0.04

5:0.07

6:0.07

7:0.11

8:0.06

9:0.08

10:0.12

11:0.10

Negative Logits

lore

-1.35

nas

-1.31

nir

-1.21

gaard

-1.20

ienne

-1.19

ega

-1.18

uru

-1.18

hello

-1.15

song

-1.15

kHz

-1.14

POSITIVE LOGITS

 policing

1.45

 protecting

1.38

 securing

1.25

 bystanders

1.24

MFT

1.23

 shielding

1.21

angering

1.18

ifully

1.18

 safeguard

1.17

illance

1.17

Activations Density 0.023%

terms related to safeguarding and security

No Comments

No Known Activations