INDEX

Explanations

information related to illegal drugs, specifically heroin, including mentions of its effects, purity, associated substances, medical use, and societal impact

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

words associated with heroin and its effects or related issues

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

0:0.07

1:0.05

2:0.19

3:0.06

4:0.04

5:0.08

6:0.03

7:0.04

8:0.06

9:0.22

10:0.08

11:0.04

Negative Logits

BUG

-1.58

iHUD

-1.58

EMBER

-1.51

quickShipAvailable

-1.44

Unity

-1.43

AIR

-1.40

THING

-1.37

ِ

-1.37

EEK

-1.37

URI

-1.36

POSITIVE LOGITS

 contamin

1.38

 needles

1.34

olate

1.33

 dispensaries

1.29

 dispensary

1.24

 blight

1.22

 mills

1.22

 Archdemon

1.22

enc

1.20

aundering

1.19

Activations Density 0.021%

information related to illegal drugs, specifically heroin, including mentions of its effects, purity, associated substances, medical use, and societal impact

words associated with heroin and its effects or related issues

No Comments

No Known Activations