INDEX

Explanations

phrases related to the significance and importance of issues

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_L5_Hcat_z_lr1.20e-03_l11.00e+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.5.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

0:0.02

1:0.02

2:0.12

3:0.11

4:0.09

5:0.03

6:0.05

7:0.26

8:0.03

9:0.04

10:0.07

11:0.12

Negative Logits

redited

-1.48

BuyableInstoreAndOnline

-1.44

 guiActiveUn

-1.36

renheit

-1.28

RANT

-1.24

Ele

-1.21

ukemia

-1.20

 inexpl

-1.18

apor

-1.16

uberty

-1.15

POSITIVE LOGITS

eters

1.50

 Beware

1.46

 wedge

1.39

 beware

1.37

 vigilance

1.36

 calibr

1.31

 rethink

1.30

 Secure

1.29

 Ensure

1.29

 affordability

1.29

Activations Density 0.016%

phrases related to the significance and importance of issues

No Comments

No Known Activations