INDEX

Explanations

references to user-related data and interactions

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

0:0.02

1:0.02

2:0.06

3:0.05

4:0.05

5:0.04

6:0.31

7:0.09

8:0.04

9:0.05

10:0.14

11:0.07

Negative Logits

 Clyde

-1.51

minster

-1.40

enegger

-1.33

ioxide

-1.31

龍契士

-1.31

Wad

-1.30

 Roose

-1.29

 Straight

-1.29

 funeral

-1.27

 Queens

-1.26

POSITIVE LOGITS

intent

1.62

upload

1.60

Internet

1.50

 browsing

1.46

��

1.43

conservancy

1.41

umbn

1.37

download

1.31

boot

1.30

elected

1.30

Activations Density 0.001%

references to user-related data and interactions

No Comments

No Known Activations