INDEX

Explanations

phrases related to actions and the presence of others in various contexts

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

0:0.03

1:0.01

2:0.16

3:0.06

4:0.28

5:0.02

6:0.05

7:0.08

8:0.05

9:0.03

10:0.10

11:0.08

Negative Logits

 ourselves

-1.46

 populations

-1.42

aepernick

-1.42

 population

-1.34

cdn

-1.30

 humankind

-1.29

 Ceres

-1.29

 myself

-1.27

 oceans

-1.26

 plurality

-1.26

POSITIVE LOGITS

Reloaded

1.52

 Chou

1.43

terday

1.42

Secret

1.40

Terror

1.38

OHN

1.37

 angrily

1.36

)}

1.35

ais

1.35

Laughs

1.34

Activations Density 0.095%

phrases related to actions and the presence of others in various contexts

No Comments

No Known Activations