INDEX

Explanations

terms related to security and authority

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_L7_Hcat_z_lr1.20e-03_l11.10e+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.7.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

0:0.05

1:0.11

2:0.07

3:0.03

4:0.02

5:0.07

6:0.07

7:0.02

8:0.03

9:0.03

10:0.43

11:0.02

Negative Logits

 Soda

-2.37

 Synd

-2.23

��

-2.20

etus

-2.15

BI

-2.10

single

-2.08

nder

-2.05

 Startup

-2.03

BIT

-2.01

POSITIVE LOGITS

 Guard

4.10

guard

3.89

 guard

3.79

 Guards

3.67

guards

3.64

Guard

3.28

 guards

3.27

 Guardians

2.84

 guarding

2.83

 militia

2.74

Activations Density 0.001%

terms related to security and authority

No Comments

No Known Activations