Lightnews — Scholar-powered news

Lily Xu @lilyxu.bsky.social · 7d

Link to application: apply.interfolio.com/172496

Applications will be evaluated on a rolling basis, with priority given to those received before October 20.

Lily Xu @lilyxu.bsky.social · 7d

Columbia is hiring a tenure-track asst. prof. in computational ecology and evolution!

If you use or develop computational tools—AI, data science, stats—to deepen our understanding of eco/evo, evaluate nature-based solutions, or inform and manage impacts on human wellbeing, please apply!

1 2 5

Lily Xu @lilyxu.bsky.social · 9d

Very happy to be part of this new paper brilliantly led by @calebscoville.bsky.social on new opportunities for stakeholder participation with dynamic management compared to static management of natural resources! 🌳🌏🙋

Caleb Scoville @calebscoville.bsky.social · 11d

New article out in Geo: Geography and Environment with an interdisciplinary dream team of coauthors: “From maps to models: Participation and contestability in the dynamic management of natural resources.”
doi.org/10.1002/geo2... (open access)

How does stakeholder participation in natural resource management change when conservation rules are grounded in near real-time data? Recent technological advances have increased the feasibility of the ‘dynamic management’ of natural resources, which promises to align the spatiotemporal scales of management with ecological variability and resource use. Drawing on Kelty's (2020) concept of ‘contributory autonomy’, this article offers a critical comparison of how participation is conceived of in the more established context of static conservation areas and planning versus the emergent field of dynamic management. A systematic review of the dynamic ocean management literature reveals a varied, but shallow engagement with the topic of stakeholder participation in that context. Whereas static management regimes are governed by relatively intuitive and contestable maps, dynamic management is governed by models and data flows. Overall, the decision-making stakeholder of participatory mapping processes under static management is displaced by the stakeholder conceived as an ‘end-user’ of a dynamic management product and consultant in its design. Yet, these shifts also open up potential points of contestation, which may pattern the future theory and practice of participation in dynamic management: counterdata, countermodelling and data chokepoints. Beyond the empirical focus on oceans, this article contributes to broader conversations about the political stakes of environmental data, and algorithmic and artificial intelligence-driven natural resource conservation by considering how possibilities for participation are foreclosed, enabled and reconstituted by new spatiotemporal and technological conditions.

1 2 7

Reposted by Lily Xu

Rachit Dubey @rachitdubey.bsky.social · Sep 3

My lab at UCLA is hiring 1-2 PhD students this cycle!

Join us to work at the intersection of cognitive science and AI applied to pressing societal challenges like climate change.

More info about me: rachit-dubey.github.io

My lab: ucla-cocopol.github.io

Please help repost/spread the word!

2 20 39

Reposted by Lily Xu

Sara Beery @sarameghanbeery.bsky.social · May 12

The call for applications has just been released for #CV4Ecology2026!! This three-week intensive program trains ecologists and conservation practitioners to develop their own AI tools for their own data.

When: Jan 12-30, 2026
Where: SCBI @smconservation.bsky.social

1 19 30

Lily Xu @lilyxu.bsky.social · May 15

Dave Thau, @sarameghanbeery.bsky.social, and I are organizing biodiversity webinar talks for @itu.int AI for Good.

Next week, I'll chat with Iadine Chades (Monash) on the promise and impact of decision AI for biodiversity.

Please share questions you'd love to ask Iadine or our upcoming speakers!

3 11

Lily Xu @lilyxu.bsky.social · Apr 29

Congratulations to you Rajanie and the team! ☀️

1 1

Lily Xu @lilyxu.bsky.social · Apr 25

PDF of advert, for sharing: lily-x.github.io/files/invasi...

1

Lily Xu @lilyxu.bsky.social · Apr 25

I'm hiring for a machine learning data scientist & research assistant for summer 2025!

Join me on a project on invasive species management with an innovative startup doing on-the-ground removal of environmentally destructive invasive animals.

Paid, full-time w/ possibility to extend.

I'm hiring for a machine learning data scientist & research assistant for summer 2025!

Join me in working on a project on invasive species management, by predicting species risk and planning capture strategies. This work will be in partnership with an innovative startup that is working directly to capture environmentally destructive invasive species in the US.

This project will be working with a US-based conservation partner to build predictive machine learning models and design harvest strategies for the removal of invasive animal species. The primary goal is to develop usable ML models and optimization tools to inform practical environmental decision-making on the ground.

There will also be opportunities to extend this work into a research publication for a top-tier AI venue.

The project would be:
Paid, full-time internship for summer 2025
Possibility of extension beyond the summer (part-time or full-time)
Remote, but candidates should be authorized to work in the US
Supervised by Lily Xu, assistant professor at Columbia University

Ideal candidate background will include:
Strong background in CS, data science, and/or applied math
Experience developing machine learning models
Excellent coding skills in Python
Excellent writing and interpersonal communication skills
Genuine interest in conservation/sustainability
Nice to have: background in optimization methods
Nice to have: experience with GIS and geospatial data
Candidates would ideally have already completed an undergraduate degree, but exceptional undergrads will also be considered.

How to apply:
Please send a CV and 3–5 paragraphs of your background/interests via email to <lily.x@columbia.edu> with the subject line "Application: ML for invasive species management".

Applications will be reviewed on a rolling basis.

1 12 23

Lily Xu @lilyxu.bsky.social · Apr 24

Monday
Hall 4 #3

Climate Change AI workshop, where we have:
🧪 @yisongyue.bsky.social keynote on AI for Science
🐘 Iadine Chades keynote on AI for Biodiversity

Panels
🌏 Opportunities for ML for Climate Action in Asia
☀️ Data-Centric ML in Climate Applications

www.climatechange.ai/events/iclr2...

Tackling Climate Change with Machine Learning

ICLR 2025 Workshop: Tackling Climate Change with Machine Learning

www.climatechange.ai

1 3

Lily Xu @lilyxu.bsky.social · Apr 24

Sunday @ 9am
Peridot 201 & 206

Invited talk at XAI4Science workshop, where I'll be speaking on "Barriers to explainability for sustainability, and a path towards more reliable models"

xai4science.github.io

XAI4Science Workshop

xai4science.github.io

1 1

Lily Xu @lilyxu.bsky.social · Apr 24

Friday evening @ 6:30pm

@climatechangeai.bsky.social and Future Cities Laboratory (FCL) Global, Singapore-ETH Centre satellite event

I'll be a panelist on "Data-Centric Approaches to ML for Climate Action"

www.eventbrite.sg/e/ccai-x-fcl...

[CCAI x FCL Global] Sustainability x AI Networking Forum

This satellite event of ICLR 2025 brings together individuals working at the intersection of artificial intelligence and climate change.

www.eventbrite.sg

1

Lily Xu @lilyxu.bsky.social · Apr 24

Paper:
Deep RL + mixed integer programming to plan for restless bandits with combinatorial (NP-hard) constraints.

with @brwilder.bsky.social, Elias Khalil, @milindtambe-ai.bsky.social

Poster #416 on Friday @ 3–5:30pm

bsky.app/profile/lily...

Lily Xu @lilyxu.bsky.social · Apr 1

Can we use RL to plan with combinatorial constraints?

Our #ICLR2025 paper combines deep RL with mathematical programming to do so! We embed a trained Q-network into a mixed-integer program, into which we can specify NP-hard constraints.

w/ brwilder.bsky.social, Elias Khalil, Milind Tambe

Our method combines deep RL with mixed-integer programming
for sequential planning with combinatorial actions.

1 2 4

Lily Xu @lilyxu.bsky.social · Apr 24

Thrilled to be here in Singapore for #ICLR2025!

Find me:

Friday @ 3–5:30pm
Poster #416
RL with combinatorial actions

Friday @ 6:30pm
Panelist at CCAI x ETH event

Sunday @ 9am
Invited talk @ XAI4Science workshop

Monday all day
Hall 4 #3
Organizing @climatechangeai.bsky.social workshop

1 2 8

Lily Xu @lilyxu.bsky.social · Apr 16

Thrilled to be co-organizing the Tackling Climate Change with Machine Learning workshop at #ICLR2025 and bringing in these fantastic keynote speakers!

Join us April 28 (in Singapore or free livestream) to hear from @yisongyue.bsky.social on AI for science and Iadine Chades on AI for biodiversity.

Climate Change AI @climatechangeai.bsky.social · Apr 16

We're excited to announce @yisongyue.bsky.social and Iadine Chades as keynote speakers at our #ICLR2025 workshop on Apr 28! 🎉

Learn more about the workshop and our free livestream here: www.climatechange.ai/events/iclr2...

A big thanks to @mila-quebec.bsky.social for sponsoring the event.

1 5 9

Reposted by Lily Xu

Yadvinder Malhi @ymalhi.bsky.social · Apr 11

Are you interested in boldly reimagining how we can finance #naturerecovery at scale, beyond mainstream market mechanisms?
We are advertising a one-year postdoctoral researcher at
@naturerecovery.bsky.social at Oxford. Closing date April 30th
tinyurl.com/muebaf86
@oxfordgeography.bsky.social

14 19

Reposted by Lily Xu

Center for Effective Global Action @cega-uc.bsky.social · Mar 19

🌎 Join CEGA for #MeasureDev2025: Biodiversity on Land and at Sea on April 29 at The World Bank in Washington, DC!

Explore cutting-edge biodiversity measurement approaches with leading researchers & policymakers shaping conservation strategies. #datasky

Register: cega.berkeley.edu/event/measur...

Measuring Development 2025: Biodiversity on Land and at Sea

Human flourishing depends on biodiversity: the animals, plants, and other living organisms whose functions make Earth habitable and whose existence holds intrinsic value. New data and applications...

cega.berkeley.edu

2 1

Lily Xu @lilyxu.bsky.social · Apr 1

I hope to see many of you in Singapore at ICLR — find me at my poster (April 25 @ 3–5:30pm SGT), at the @climatechangeai.bsky.social workshop (April 28), or please do reach out to grab a coffee!

🇸🇬☕️🦁🏙️👩🏻‍🔬

1

Lily Xu @lilyxu.bsky.social · Apr 1

I'm also co-organizing the @climatechangeai.bsky.social workshop at #ICLR2025!

Join us April 28: climatechange.ai/events/iclr2...

We're thrilled to have keynotes by @yisongyue.bsky.social and Iadine Chades on path-defining work on AI for science and AI for conservation.

We'll be livestreaming!

Tackling Climate Change with Machine Learning

ICLR 2025 Workshop: Tackling Climate Change with Machine Learning

climatechange.ai

1 2 2

Lily Xu @lilyxu.bsky.social · Apr 1

Reinforcement learning with combinatorial actions for coupled restless bandits

Paper: arxiv.org/abs/2503.01919
Code: github.com/lily-x/combi...

Join our #ICLR2025 poster session!
Friday, Apr 25 @ 3–5:30pm SGT

Poster for Reinforcement learning with combinatorial actions for coupled restless bandits

1 2

Lily Xu @lilyxu.bsky.social · Apr 1

Our SEQUOIA method significantly improves performance above baseline approaches, which are either myopic (neglecting the sequential nature of RMABs and other MDPs) or non-combinatorial (using heuristic, iterative approaches to solve the NP-hard combinatorial constraints).

1

Lily Xu @lilyxu.bsky.social · Apr 1

Our method combines:
— DQN: deep Q-network to approximate the value function
— MILP solving: a mixed-integer program to embed combinatorial constraints over actions

The MILP solver implicitly conducts a forward pass through the neural network to compute the expected Q-value.

Diagram of our SEQUOIA method. Part 1: training, to learn a Q-function approximator; part 2: evaluation to find the action that maximizes the expected Q-value.

1

Lily Xu @lilyxu.bsky.social · Apr 1

Sequential planning with per-timestep combinatorial (NP-hard) constraints is hard.

Existing RL for combinatorial planning focus on
(a) only combinatorial state spaces, or
(b) RL as a heuristic solver for combinatorial optimization — but in single-shot (non-sequential) planning.

1

Lily Xu @lilyxu.bsky.social · Apr 1

We focus on planning for restless bandits (RMABs), a form of constrained MDPs. RMABs are used for public health planning with ARMMAN, an NGO in India.

Traditional RMABs model only knapsack "top-K" actions, which do not capture many practical constraints such as scheduling or path-planning.

Maternal health interventions in India may include scheduling, capacity, or routing constraints.

1

Lily Xu @lilyxu.bsky.social · Apr 1

Can we use RL to plan with combinatorial constraints?

Our #ICLR2025 paper combines deep RL with mathematical programming to do so! We embed a trained Q-network into a mixed-integer program, into which we can specify NP-hard constraints.

w/ brwilder.bsky.social, Elias Khalil, Milind Tambe

1 1 6