INDEX

Explanations

references to research methodology and study results

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

0:0.09

1:0.02

2:0.06

3:0.13

4:0.06

5:0.03

6:0.06

7:0.19

8:0.05

9:0.05

10:0.09

11:0.13

Negative Logits

 airs

-1.60

 Wander

-1.30

 laure

-1.25

milo

-1.24

 dearly

-1.22

 fert

-1.20

 Xeon

-1.20

"},

-1.19

iven

-1.17

iban

-1.16

POSITIVE LOGITS

・

1.34

Discussion

1.30

Dest

1.28

velt

1.27

idable

1.24

enaries

1.24

Decre

1.24

 Slaughter

1.22

ッ

1.22

alyses

1.22

Activations Density 0.004%

references to research methodology and study results

No Comments

No Known Activations