INDEX

Explanations

references to physical objects and components

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

0:0.45

1:0.02

2:0.03

3:0.03

4:0.10

5:0.05

6:0.04

7:0.06

8:0.03

9:0.02

10:0.09

11:0.02

Negative Logits

Ranked

-1.41

 Competitive

-1.38

 Instant

-1.35

 Hunger

-1.30

 Rebirth

-1.27

nesota

-1.24

 Commun

-1.20

 casual

-1.19

 mainstream

-1.19

 Multiplayer

-1.18

POSITIVE LOGITS

pole

2.21

 cleaners

1.55

 blades

1.52

ridges

1.48

holes

1.46

addle

1.46

iffe

1.44

riers

1.43

boards

1.42

 lengths

1.42

Activations Density 0.716%

references to physical objects and components

No Comments

No Known Activations