INDEX

Explanations

references to ranking or listing items

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

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

0:0.01

1:0.01

2:0.06

3:0.06

4:0.26

5:0.03

6:0.08

7:0.20

8:0.03

9:0.03

10:0.08

11:0.10

Negative Logits

licensed

-1.29

 interacted

-1.28

pressed

-1.27

tm

-1.26

 located

-1.24

operated

-1.23

ococ

-1.23

ocated

-1.22

 volunteers

-1.20

mu

-1.19

POSITIVE LOGITS

 ranks

1.99

ynski

1.51

 shelves

1.50

 fray

1.49

ahime

1.46

 charts

1.44

tein

1.36

 classroom

1.35

 rankings

1.32

 benchmarks

1.32

Activations Density 0.005%

references to ranking or listing items

No Comments

No Known Activations