Siyuan Guo
@syguoML
Followers
1,084
Following
188
Media
11
Statuses
170
PhD Student in ML with @Cambridge_Uni @MPI_IS . Working on Causal Exchangeability and AI4Science. Research Scientist Intern @AIatMeta . 🇨🇳
Joined May 2021
Don't wanna be here?
Send us removal request.
Explore trending content on Musk Viewer
hobi
• 201609 Tweets
Tuchel
• 192825 Tweets
FY RECAP BLANK SS2EP4
• 164640 Tweets
JIN IS COMING
• 157930 Tweets
HAPPY TRACKLIST
• 156442 Tweets
LINGLING MYCHERIE AMOUR EP8
• 124895 Tweets
カズラドロップ
• 66544 Tweets
TXT ANGEL CONCEPT PHOTO & CLIP
• 50426 Tweets
Taka
• 46231 Tweets
Another Level
• 44607 Tweets
$STORM
• 21127 Tweets
#JO1ANNX
• 16082 Tweets
ベイスターズ
• 13996 Tweets
色違いメロエッタ
• 12924 Tweets
#القمه_الخليجيه_الاوروبيه
• 11955 Tweets
STRAY KIDS COME TO PLAY
• 10576 Tweets
Pinned Tweet
New preprint: Do Finetti w/
@zcccucla
,
@Carthica
,
@fhuszar
,
@bschoelkopf
and me.
Do Finetti provides a do-calculus foundation for exchangeable data following the independent causal mechanism (ICM) principle + a causal Pólya urn model to show how
5
23
131
Want to do a PhD in ML, consider applying to the fully-funded Cambridge - Tübingen PhD Fellowship.
From personal experience supervised by
@fhuszar
and
@bschoelkopf
, this is the best PhD out there!
Deadline: noon, Dec 5th.
8
38
220
Super excited
#neurips2024
Oral for Do Finetti!
See you in 🇨🇦 to discuss exchangeability, causality and Nature intelligence!
New preprint: Do Finetti w/
@zcccucla
,
@Carthica
,
@fhuszar
,
@bschoelkopf
and me.
Do Finetti provides a do-calculus foundation for exchangeable data following the independent causal mechanism (ICM) principle + a causal Pólya urn model to show how
5
23
131
6
8
128
Generalization is more than distribution shifts.
work w/
@WildbergerJonas
,
@bschoelkopf
,
We proposed out-of-variable (OOV) generalization to study knowledge transfer with partial observability. We showed learning from the residual uncovers unobserved
4
16
119
Will be talking about our recent work Causal de Finetti () in London on November 4th, 12pm midday UK time at
@uclcsml
.
Feel free to reach out if you are around and have a chat about Causal ML!
Details can be found here:
3
15
111
I will be attending
#NeurIPS2023
and present causal de Finetti @ 14 Dec 10:45 - 12:45. DM if you want to chat about causality, AI for science and RAG!
We provably show that properties of input data affect models' downstream properties e.g. compositionality, disentanglement.
preprint by
@syguoML
w/ Viktor Tóth,
@bschoelkopf
&me
New causal de Finetti theorems provide a probabilistic foundation for the independent causal mechanisms (ICM) principle + a principled way to infer invariant causal structure from exchangeable data
🧵👇
3
24
158
2
8
76
Causal de Finetti had made it to
#NeurIPS2023
! Shout out to my amazing supervisors and collaborators. Very excited for the field of causal exchangeability.
Meet you all in New Orleans!
preprint by
@syguoML
w/ Viktor Tóth,
@bschoelkopf
&me
New causal de Finetti theorems provide a probabilistic foundation for the independent causal mechanisms (ICM) principle + a principled way to infer invariant causal structure from exchangeable data
🧵👇
3
24
158
1
3
63
Beyond excitement to receive this award - thank you for people who believed in me and recognised my work! Thank you to the organisers for creating this award and your efforts in promoting gender equality in
#research
community! Particularly, my advisors
@bschoelkopf
and
@fhuszar
4
4
49
For people of interest, please see Twitter summaries on:
Do Finetti (2024):
causal de Finetti (2023):
Recommending chronological reading order :)
Thank you
@Dagophile
for dropping everything and giving us a summary of the
@syguoML
etal paper. Following your example, I'm going to drop everything and read your summary.
1
4
10
1
4
22
Thank you
@AleksanderMolak
for the nice intro! Indeed, we view data heterogeneity, especially the regime of exchangeable but not i.i.d. data, as an opportunity for causality.
Our work shows exchangeable data has extra conditional independence structures that i.i.d. data lacks -
What if we could discover the true causal structure from observational data?
Too good to be true?
1/n
#causaltwitter
#causality
#machinelearning
#causaldiscovery
3
44
153
1
2
24
Started to think more about the topic of causal digital twins by organizing this small workshop at ELLIS unconference with Andrei,
@lawrennd
and
@bschoelkopf
. What is causal? And why digital twins? Looking forward to future works coming from this community :)
It was the first
#unconference
for
#ELLISforEurope
: Scientists from our network met in Spain to discuss the latest cutting-edge
#ML
research during a fully participant-driven event which resulted in many ideas for future collaborations!
#AI
0
3
17
0
1
18
Good morning, beautiful Arenzano! Excited to start the day with ELLIS Theory Workshop!
#malga
#ellisforeurope
0
0
16
In Vienna
#ICLR2024
to present OOV Generalization for discriminative models.
🗓️Thu 9 May 4:30 p.m.
📍Hall B
#190
Looking forward to chat about Causality & AI for Science!! Drop me a message :D
Generalization is more than distribution shifts.
work w/
@WildbergerJonas
,
@bschoelkopf
,
We proposed out-of-variable (OOV) generalization to study knowledge transfer with partial observability. We showed learning from the residual uncovers unobserved
4
16
119
0
0
13
Congratulations to ELLIS! And a great fortune for students from European and the wider world 🌍
@ELLISforEurope
@ELLISInst_Tue
@MPI_IS
@Cyber_Valley
Fascinating visitors at our institute this week, with a CIFAR workshop & a scientific symposium for the upcoming ELLIS institute opening (1/2)
3
20
159
0
0
13
Don’t miss when Patrik gives a talk👇
I am grateful for the opportunity to present our work on Identifiable Exchangeable Mechanisms at the
@Mila_Quebec
tea talks this Friday at 10:30 EDT.
Joint work with
@syguoML
,
@fhuszar
,
@wielandbr
,
@bschoelkopf
.
Preprint: (v2 is up tomorrow)
4
9
28
0
0
9
A fun project bridging
#causal
and
#fairness
: the key is that one needs to consider the action space available to achieve pragmatic fair outcomes.
Joint work with
@gultchin
, Alan Malek,
@csilviavr
, Ricardo Silva from
@DeepMind
@ai_ucl
Paper:
I'm at
#NeurIPS2022
! Looking forward to in-person chats.
Will be presenting new work (Pragmatic Fairness) at the AFCP workshop on Saturday Dec 3rd.
Till then, which
#CausaML
and
#TrustworthyML
works should I check out? Looking for recs!
0
12
69
0
3
8
This Wed. 10:30 ET (22nd Nov.) at the CDG we are honored to have
@zcccucla
discussing their
@NeurIPSConf
paper: Causal Inference w Non-IID Data using Linear Graphical Models ()
All info & links @ 🌿
cc:
@yudapearl
@Carthica
0
4
17
0
0
8
DAY 0 at NeurIPS: cancels hotel reservation at last minute after flying across half the globe. Now searching somewhere to live... true NeurIPS experience +1
2
0
7
Thrilled to share our preprint - Causal de Finetti w/ Viktor Tóth
@fhuszar
@bschoelkopf
Check it out!
preprint by
@syguoML
w/ Viktor Tóth,
@bschoelkopf
&me
New causal de Finetti theorems provide a probabilistic foundation for the independent causal mechanisms (ICM) principle + a principled way to infer invariant causal structure from exchangeable data
🧵👇
3
24
158
0
1
7
@zcccucla
@Carthica
@fhuszar
@bschoelkopf
[3/n] We proved a generalized truncated factorization that shows causal effects are identifiable in ICM exchangeable processes. Traditional truncated factorization is a special case here, just as how i.i.d. is a special case of exchangeable.
1
1
6
Very exciting to see Goutham Rajendran,
@rpatrik96
, Pradeep Ravi Kumar,
@wielandbr
connecting causal de Finetti to dynamical systems.
We believe humans discover causal relationships through exploration in time, and we need to understand how.
There is a bonus insight!
Hidden Markov Models (LTI systems are HMMs) are used to represent causal relationships. We show that learning HMMs follows the blueprint of the Causal de Finetti (CdF) theorem by
@syguoML
, Viktor Tóth,
@fhuszar
,
@bschoelkopf
.
1
0
0
0
1
5
[5/n] We introduce the causal Pólya urn model:
Suppose you are an observer in front of a black-box. You observe Xn, Yn at time step n. If Xn = 1, you put a red ball in the left-hand side of the black box, and else a green ball. Similarly, you compute a value Zn := (1-Xn)*Yn +
1
0
5
[6/n] This game-like urn model says (Xn, Yn) satisfies the causal de Finetti conditions. Further, in the hidden (unobserved) world, our observations are indeed driven by two independent mechanisms (left and right urn). And us, observers can deduce such hidden mechanisms through
1
0
5
[9/9] We are excited to explore the complexities of non-i.i.d. data. And there is much more to do: from intervention to counterfactual, from Markovian to semi-Markovian. We see a whole world of possibilities for the general causal community in exchangeable but not i.i.d. data. I
0
0
5
[7/n] Observing more Xm = 1, Ym = 0 (m<n) means it is more likely to observe Xn = 1, Yn = 0. Conditional interventional manifests as when an intervention is performed do(Xn=0), one can deduce that it is more likely Yn = 1.
1
0
5
[8/n] Generalizing causality to an exchangeable non-i.i.d. setting does not mean less ability to perform graphical identification and effect estimation. In fact, with an unknown graph, ICM generative processes allow one to identify graph and causal effects simultaneously (cf.
1
0
4
@zcccucla
@Carthica
@fhuszar
@bschoelkopf
[2/n] Do-calculus is based on structural causal models. However, SCMs fail to characterize ICM exchangeable settings. We need a new definition of what intervention means.
1
0
4
[4/n] Causal effects in ICM exchangeable processes have non-trivial conditional interventional distributions. This property does not exist in i.i.d. data. See Block B in Figure 1 and we demonstrate this in the causal Pólya urn model below.
2
0
4
@WildbergerJonas
@bschoelkopf
Take a simple example: Markov factorization is composed of causal Markov kernels though one does not need to observe and measure all variables of interest to recover the joint distribution. Rather, observing only variables in the Markov kernel per environment is sufficient to
1
0
3
More of a complaint on Thanks everyone for offering to help. all sorted
0
0
2
@VStimper
@ArnaudDoucet1
@pl219_Cambridge
@jmhernandez233
@bschoelkopf
Congratulations Dr. Stimper! All the best for your next journey. Gonna miss the old days of office chats! ❤️
0
0
1
@sangmichaelxie
Very like your work on in context learning and exchangeablity. I am working on exchangeablity with causality. would like to chat!
0
0
2
Day 1 @ NeurIPS 2023:
@lindensli
's talk is fantastic, do check out the slides once available! It explains with such clarity the tools used daily for LLM inference speed-up and its underlying principles, e.g., tensor parallelism, vllm.
Highlights are:
1. LLM inference on GPU
I’ll be giving a talk tomorrow at NeurIPS about the fundamentals of LLM inference. The talk will start by developing a first principles, systems-approach to reasoning about the inference workload and conclude with a survey of the current state of the art. Some concepts covered:
13
21
156
0
0
1
Great insights by
@DAGophile
on Do Finetti and causal de Finetti series!
I dropped everything to read
@syguoML
@zcccucla
@Carthica
@fhuszar
&
@bschoelkopf
's paper (+the related
@NeurIPSConf
'23 paper), which has dramatic implications for understanding (& generalizing) the relationship between causation and probability
a philosopher's 🧵
2
8
36
0
0
2
@laurence_ai
@fhuszar
@bschoelkopf
intuitively this is possible because our CI encodes P_{Y|X} is independent of P_X!
1
0
1
@laurence_ai
@fhuszar
@bschoelkopf
Great insight! This is exactly how we looked at it. Our theorem formally shows that if exchangeable and satisfy additional CI then we can always represent our data like the figure (theta and psi are independent).
1
1
1
@rpatrik96
Thank you for being an amazing collaborator - it is my great pleasure to work on interesting research together!
0
0
1
We study a seemingly impossible toy example in this work where Y is the effect of three independent causes under ANM. Our goal is to learn more about the predictive function in the target domain using source domain and marginal information about target covariates only.
1
0
1
The key insight is to learn from the residual distribution. We show the moments of the residual distribution are composed of the moments of the target covariates and the partial derivative with respect to the target covariate (hidden to us in the source domain).
1
0
1
@VStimper
@IsomorphicLabs
Congratulations Vincent! Very looking forward to your next journey - gonna be a blast as always ;)
0
0
1
@analisereal
Thanks for your question! This work studies causal effect estimation under different data-generating processes. In general, causal effect is identifiable in Markovian models for i.i.d. but here Fig. 3 aims to show causal effect has extra properties in exchangeable but not i.i.d.
1
0
1
@JKugelgen
@jmhernandez233
@RavikumarPrad
Congratulations Julius! Going to keep following your exciting research work!
0
0
1
@LuciaCKun
@MPI_IS
@laion_ai
It was great to meet you! 😃 Lovely to have interesting discussions together!
1
0
1
@ZhijingJin
Yes I am experiencing the same. Due to cybersecurity issues on Microsoft IT systems. Now trying hard to rebook. Fingers crossed for you!
0
0
1
1
0
1