INDEX

Explanations

mentions of authorities or official entities in contexts related to safety or threats

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

0:0.06

1:0.03

2:0.07

3:0.03

4:0.04

5:0.04

6:0.22

7:0.03

8:0.07

9:0.29

10:0.02

11:0.05

Negative Logits

phy

-4.32

Pod

-4.08

ubby

-3.89

bom

-3.88

NBC

-3.73

�

-3.62

pod

-3.56

BUR

-3.54

Pod

-3.54

CPC

-3.54

POSITIVE LOGITS

Guard

9.67

 Guardian

7.63

 Guard

7.52

guard

7.47

 Guardians

7.37

 Guards

7.20

 guard

7.12

guards

6.94

 guards

6.69

 guardian

5.81

Activations Density 0.006%

mentions of authorities or official entities in contexts related to safety or threats

No Comments

No Known Activations