INDEX

Explanations

negative or critical assessments of established systems or practices

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

0:0.02

1:0.03

2:0.06

3:0.07

4:0.05

5:0.04

6:0.43

7:0.06

8:0.05

9:0.05

10:0.05

11:0.04

Negative Logits

oglu

-1.43

wagen

-1.31

 Reloaded

-1.30

 Stre

-1.20

 Cooke

-1.19

 Victory

-1.19

eday

-1.18

 Fenrir

-1.17

 stakes

-1.15

itivity

-1.13

POSITIVE LOGITS

ascript

1.46

ilitarian

1.42

existent

1.35

ussed

1.34

SpaceEngineers

1.34

core

1.32

med

1.30

profit

1.29

represented

1.28

smoking

1.27

Activations Density 0.030%

negative or critical assessments of established systems or practices

No Comments

No Known Activations