INDEX

Explanations

instances of the word "hide" and variations like "hiding" in a text

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

instances of hiding or concealing information or identity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-att-kk

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L2_Hcat_z_lr1.20e-03_l11.00e+00_ds24576_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v4.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

float32

Hook Name

blocks.2.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

0:0.06

1:0.02

2:0.15

3:0.08

4:0.05

5:0.08

6:0.03

7:0.03

8:0.23

9:0.14

10:0.05

11:0.03

Negative Logits

 onwards

-1.36

 Kinnikuman

-1.26

Pwr

-1.24

 racer

-1.21

 brewers

-1.19

NRS

-1.17

 Prix

-1.17

liest

-1.17

 Vish

-1.17

eller

-1.16

POSITIVE LOGITS

artifacts

1.39

 secrets

1.37

yet

1.30

jah

1.27

aths

1.27

 Secrets

1.26

thin

1.24

jri

1.23

 illusions

1.21

lvl

1.19

Activations Density 0.023%

instances of the word "hide" and variations like "hiding" in a text

instances of hiding or concealing information or identity

No Comments

No Known Activations