INDEX

Explanations

phrases indicating actions taken by individuals or organizations

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.03

1:0.02

2:0.11

3:0.13

4:0.11

5:0.03

6:0.03

7:0.18

8:0.06

9:0.05

10:0.11

11:0.11

Negative Logits

iciency

-1.65

payer

-1.59

hya

-1.52

obyl

-1.51

akespeare

-1.48

iors

-1.47

gdala

-1.44

ynes

-1.44

knowledge

-1.41

vironments

-1.40

POSITIVE LOGITS

Finish

1.47

 scheduling

1.44

Primary

1.43

 captains

1.37

 reigning

1.37

 Schmidt

1.34

 Manz

1.29

 Ridley

1.28

 midway

1.27

 Lone

1.27

Activations Density 0.001%

phrases indicating actions taken by individuals or organizations

No Comments

No Known Activations