INDEX

Explanations

words related to ranking or ordering systems

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

1:0.05

2:0.17

3:0.06

4:0.03

5:0.05

6:0.18

7:0.04

8:0.03

9:0.20

10:0.05

11:0.03

Negative Logits

 Chomsky

-3.82

 Willis

-3.80

 Dempsey

-3.75

 Michaels

-3.71

 Certification

-3.68

 Doctors

-3.67

 dict

-3.53

 Walter

-3.50

 Clayton

-3.48

Wal

-3.46

POSITIVE LOGITS

 attracted

7.19

 attraction

7.13

 attract

7.02

 attracts

6.82

 attracting

6.53

 attractions

5.79

 lure

5.38

 lured

5.05

 attractiveness

4.45

isexual

4.37

Activations Density 0.001%

words related to ranking or ordering systems

No Comments

No Known Activations