LIWAIWAI
beneath and beyond,
now and next
Independent ASEAN AI Institute for research, engineering, assurance, and deployment.
LIWAIWAI
An institute, not a vendor. We hold no commercial stake in the systems we assess.
An independent ASEAN institute researching artificial intelligence, its risks, its opportunities and its impact on society.
liwaiwai brings together research, engineering, assurance, deployment, community, advocacy, reassurance and engagement to help the region understand and navigate AI.
We work across technical, policy and societal dimensions, serving researchers, governments, institutions, industry, practitioners, civil society and the wider public.
Our work is open by default. We publish methods and findings, partner with institutions across the region, and treat independence as a practice rather than a claim.
4
READ dimensions
4
CARE dimensions
10
ASEAN states
Open
Independent
READ
Understanding AI, evaluating its risks and developing practical approaches to responsible deployment.
READ → UNDERSTAND
Research · Engineering · Assurance · Deployment
Four connected dimensions of how liwaiwai studies and builds with AI.
Together they form a continuous loop: study, build, test and apply, then feed what we learn back into research.
- 01
Research
Independent research into AI, governance, safety, technology and society. We monitor capabilities, evaluate risks and publish open findings for the region.
- 02
Engineering
Technical experimentation, tools, benchmarks and systems supporting the responsible understanding of AI. We build open evaluation tooling and test harnesses others can reuse.
- 03
Assurance
Evaluation, assessment, testing, safety and reliability. We assess systems before and after release so organisations and the public can trust what is deployed.
- 04
Deployment
Understanding how AI can be implemented responsibly in organisations and institutions. We translate research into practical guidance for adopting AI across sectors.
CARE
Connecting people, institutions and communities to navigate the opportunities and challenges of AI.
CARE → RESPOND
Community · Advocacy · Reassurance · Engagement
Four connected dimensions of how liwaiwai works with people, not just with systems.
CARE is how liwaiwai stays accountable to the people AI affects, not only to the organisations that build or buy it.
About CARE at liwaiwai →- 01
Community
Connecting researchers, practitioners, institutions and wider society. We convene people around shared questions across the region.
- 02
Advocacy
Supporting informed discussion and responsible approaches to AI. We champion evidence-based, proportionate governance across ASEAN.
- 03
Reassurance
Helping people understand AI risks and opportunities through evidence and clear communication, translating technical risk into plain language.
- 04
Engagement
Events, sessions, consultations, presentations, calls and public dialogue that keep the conversation open and accountable.
DEVELOPING
ASEAN AI Observatory
Tracking how artificial intelligence is developing, deployed and governed across Southeast Asia. A regional view of capability, policy and impact, built for researchers, institutions and the public.
Explore the ObservatoryResearch
Selected work and findings from across liwaiwai's research programmes.
- 02 Oct 2026
Transferable Graph Metanetworks
AI Research & Preprints · arXiv - Statistics (stat.ML) · Yuxin Ma, Adir Dayan, Yam Eitan, Haggai Maron, Soledad Villar
arXiv:2610.00420v1 Announce Type: new Abstract: A weight space network (or metanetwork) takes the weights of another neural network as input and predicts properties of it. Most pri…
- 02 Oct 2026
Beyond Lie Algebras: Lie-Wedge Stratification and Pure-State Stabilizability in Single-Channel Qubit Control
Science & Mathematics · arXiv - Quantum Physics · Yibin Wang
arXiv:2610.00046v1 Announce Type: new Abstract: Lie algebras describe how control generators combine, but in open systems they erase the distinction between reversible control and …
- 02 Oct 2026
ChainLoRA: Geometry-Preserving Task Vector Merging for Continual Learning in LLMs
AI Research & Preprints · arXiv - Statistics (stat.ML) · Hang Yin, Haozhe Wang, Yuhua Luo, Zhangqi Pan, Xiaoxing Wang, Junchi Yan
arXiv:2610.00431v1 Announce Type: new Abstract: Continual parameter-efficient fine-tuning for large language models (LLMs) must balance retention of previously acquired knowledge, …
ASEAN AI Brief
AI developments that matter to ASEAN. A continuing synthesis of what is happening in AI and what it means for the region.
National Economic Power Industry War Conference
Please join ITIF for an important policy conference on what U.S. policymakers must do to prevent America from suffering a catastrophic defea…
National Economic Power Industry War Conference
Please join ITIF for an important policy conference on what U.S. policymakers must do to prevent America from suffering a catastrophic defea…
National Power Industry War Conference
Please join ITIF for an important policy conference on what U.S. policymakers must do to prevent America from suffering a catastrophic defea…
National Economic Power Industry War Conference
Please join ITIF for an important policy conference on what U.S. policymakers must do to prevent America from suffering a catastrophic defea…
Events
Sessions, consultations and public dialogue across the region, from closed roundtables to open public sessions.
Roundtable
Closed, moderated discussions among peers on a shared question.
Research seminar
Public presentation and critique of new findings and methods.
Policy dialogue
Cross-sector conversation between officials, industry and researchers.
Workshop
Hands-on sessions for practitioners building or assessing AI systems.
Technical demonstration
Live walkthroughs of evaluation tooling, benchmarks and guardrails.
Public session
Open sessions that translate AI risk and opportunity for wider audiences.
The events calendar is being finalised.
Upcoming sessions, consultations and public dialogues will be listed here with dates and registration. To host or propose a session with liwaiwai, get in touch.
Propose a sessionPublications
Research outputs, frameworks and findings published by liwaiwai, from working papers to policy briefs.
Research
Peer-style studies of capability, risk and impact.
Report
In-depth findings from evaluation and assurance work.
Policy brief
Short, actionable notes for officials and regulators.
Working paper
Open early drafts inviting community review.
AI brief
Regular synthesis of what is happening in AI and what it means.
Commentary
Position pieces on developments across the region.
Publications will appear here.
As research outputs are published they will be listed with title, type, date and author. In the meantime, browse the live research feed for ongoing work.
Explore ResearchENGAGEMENT
Work with liwaiwai
liwaiwai works with researchers, governments, institutions, organisations, practitioners and community groups across ASEAN to understand and navigate artificial intelligence.
READ / CARE
Beneath and beyond, now and next
Independent AI research, engineering, assurance and engagement for ASEAN.