INDEX

Explanations

references to awards and their criteria or history

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

0:0.31

1:0.02

2:0.12

3:0.09

4:0.04

5:0.07

6:0.04

7:0.04

8:0.07

9:0.03

10:0.08

11:0.04

Negative Logits

 plumbing

-2.60

 backups

-2.52

 panic

-2.48

ideshow

-2.44

 debugging

-2.42

DSL

-2.38

 slowing

-2.31

 playbook

-2.31

 meltdown

-2.30

 sluggish

-2.30

POSITIVE LOGITS

 awarded

4.13

 awarding

3.56

 laure

3.34

 awards

3.34

 honoured

3.33

 award

3.32

 bestowed

3.29

 recipient

3.25

 honour

3.23

 recipients

3.18

Activations Density 0.009%

references to awards and their criteria or history

No Comments

No Known Activations