New post request: AAAI27 Workshop: Measuring Collective Intelligence in Agentic Systems

AG
Alborz Geramifard
Sun, Oct 11, 2026 7:12 PM

---===
CALL FOR PAPERS
WMAC 2027: Third Workshop on Multi-Agent Collaboration
Measuring Collective Intelligence in Agentic Systems
AAAI-27 | Montréal, Canada | February 22–23, 2027

Website:    https://multiagents.org/2027/
Submission: https://openreview.net/group?id=AAAI.org/2027/Workshop/WMAC
Deadline:  November 20, 2026 (AoE)


---===

OVERVIEW
Modern agent deployments rarely consist of a single model. Coding
assistants delegate to sub-agents for retrieval, editing, and
verification; research prototypes organize dozens of agents into
societies that negotiate and trade; and most real systems keep humans
in the loop. How well these pieces work together decides how well the
system works.

Our evaluation tools have not kept up. Standard benchmarks score one
model on one task and say little about coordination overhead,
specialization, communication efficiency, or whether adding another
agent helps at all. The first two WMAC editions asked how to make
LLM-based agents collaborate. One question kept coming back and stayed
open: how do we know a group of agents is actually working well
together? This edition takes that question as its topic.

We want to make progress on three questions:

  1. WHAT is worth measuring in a multi-agent or sub-agent system,
    beyond end-task success?
  2. HOW can it be measured reproducibly and comparably across systems?
  3. WHEN do collectives outperform individuals, and when do they merely
    add cost?

We welcome contributions from machine learning, multi-agent systems,
reinforcement learning, economics, and cognitive science.

TOPICS OF INTEREST
Submissions on the following topics are welcome, but not limited to them:

  • Metrics and evaluation methodologies: new metrics for coordination
    quality, communication efficiency, specialization, and robustness in
    agent teams; extending single-agent frameworks (e.g., AgentBoard) and
    multi-agent arenas (e.g., LLMArena) to measure collective behavior.
  • Benchmarks, environments, and simulations: controlled environments
    where collective behavior can be studied reproducibly, including
    negotiation and social settings; benchmarks that score collaboration
    and competition directly.
  • Foundations and theories: formal definitions and scaling laws for
    collective intelligence, with connections to biological and economic
    systems.
  • Hierarchical and sub-agent architectures: measurement questions
    specific to planner–executor and manager–worker structures, where
    performance is attributable to no single component.
  • Human–agent collectives: evaluation of mixed human–AI teams and
    collective decision making.
  • Applications: case studies from software engineering, scientific
    discovery, robotics, healthcare, and economics where collective
    performance was measured, successfully or not.

SUBMISSION GUIDELINES

  • Full papers: up to 8 pages. Original research on measuring collective
    intelligence in agentic systems.
  • Short, demo, and position papers: up to 4 pages. Early results, system
    demos, case studies, and position statements.
  • Page limits exclude references and appendices. Please use the AAAI-27
    author kit.
  • Reviewing is double-blind on OpenReview, with three reviews per paper.
    Please anonymize your submission.
  • The workshop is non-archival. Presenting at WMAC does not prevent
    publication elsewhere, and work under review or recently published at
    other venues is welcome.
  • Accepted papers are presented as oral talks or posters. Every poster
    gets a short lightning pitch.
  • A Best Student Paper Award will be given to work led by a master's or
    PhD student.
  • Industry demos of deployed multi-agent systems are welcome. Presenters
    should prepare a recorded fallback.

Submit at: https://openreview.net/group?id=AAAI.org/2027/Workshop/WMAC

IMPORTANT DATES
Paper submission deadline:    November 20, 2026 (AoE)
Notification of acceptance:  December 2, 2026
Workshop:                    February 22–23, 2027, Montréal

Dates follow the schedule set by the AAAI-27 workshop chairs and may
change. Check the website for updates.

PROGRAM
A full day built for interaction: four invited talks, contributed oral
and lightning talks, a poster session, breakout discussions in small
groups, live industry demos, and a closing panel.

Invited participants (tentative, pending confirmation):

  • Michael Wooldridge, University of Oxford
  • Samy Bengio, Apple / EPFL
  • Mikaela Grace, Anthropic
  • Bo An, Nanyang Technological University
  • Noah Goodman, Stanford University

ORGANIZERS

  • Raphael Shu, OpenAgents (workshop chair)
  • Alborz Geramifard, LinkedIn
  • Alex Marin, Thomson Reuters Labs
  • Yusen Zhang, Columbia University
  • Asli Celikyilmaz, Microsoft / University of Washington

ABOUT THE SERIES

WMAC is the Workshop on Multi-Agent Collaboration at AAAI. WMAC 2025
(Philadelphia) received 33 submissions with 178 registered attendees;
WMAC 2026 (Singapore, AAAI-26 Bridge Program) received 42 submissions
with 360 registered attendees. Past editions: https://multiagents.org/

CONTACT
Raphael Shu (raphael@uaca.com).

We look forward to your submissions and to seeing you in Montréal!

Best,


Alborz Geramifard https://alborz-geramifard.com/

===================================================================== CALL FOR PAPERS WMAC 2027: Third Workshop on Multi-Agent Collaboration Measuring Collective Intelligence in Agentic Systems AAAI-27 | Montréal, Canada | February 22–23, 2027 Website: https://multiagents.org/2027/ Submission: https://openreview.net/group?id=AAAI.org/2027/Workshop/WMAC Deadline: November 20, 2026 (AoE) ===================================================================== OVERVIEW Modern agent deployments rarely consist of a single model. Coding assistants delegate to sub-agents for retrieval, editing, and verification; research prototypes organize dozens of agents into societies that negotiate and trade; and most real systems keep humans in the loop. How well these pieces work together decides how well the system works. Our evaluation tools have not kept up. Standard benchmarks score one model on one task and say little about coordination overhead, specialization, communication efficiency, or whether adding another agent helps at all. The first two WMAC editions asked how to make LLM-based agents collaborate. One question kept coming back and stayed open: how do we know a group of agents is actually working well together? This edition takes that question as its topic. We want to make progress on three questions: 1. WHAT is worth measuring in a multi-agent or sub-agent system, beyond end-task success? 2. HOW can it be measured reproducibly and comparably across systems? 3. WHEN do collectives outperform individuals, and when do they merely add cost? We welcome contributions from machine learning, multi-agent systems, reinforcement learning, economics, and cognitive science. TOPICS OF INTEREST Submissions on the following topics are welcome, but not limited to them: * Metrics and evaluation methodologies: new metrics for coordination quality, communication efficiency, specialization, and robustness in agent teams; extending single-agent frameworks (e.g., AgentBoard) and multi-agent arenas (e.g., LLMArena) to measure collective behavior. * Benchmarks, environments, and simulations: controlled environments where collective behavior can be studied reproducibly, including negotiation and social settings; benchmarks that score collaboration and competition directly. * Foundations and theories: formal definitions and scaling laws for collective intelligence, with connections to biological and economic systems. * Hierarchical and sub-agent architectures: measurement questions specific to planner–executor and manager–worker structures, where performance is attributable to no single component. * Human–agent collectives: evaluation of mixed human–AI teams and collective decision making. * Applications: case studies from software engineering, scientific discovery, robotics, healthcare, and economics where collective performance was measured, successfully or not. SUBMISSION GUIDELINES * Full papers: up to 8 pages. Original research on measuring collective intelligence in agentic systems. * Short, demo, and position papers: up to 4 pages. Early results, system demos, case studies, and position statements. * Page limits exclude references and appendices. Please use the AAAI-27 author kit. * Reviewing is double-blind on OpenReview, with three reviews per paper. Please anonymize your submission. * The workshop is non-archival. Presenting at WMAC does not prevent publication elsewhere, and work under review or recently published at other venues is welcome. * Accepted papers are presented as oral talks or posters. Every poster gets a short lightning pitch. * A Best Student Paper Award will be given to work led by a master's or PhD student. * Industry demos of deployed multi-agent systems are welcome. Presenters should prepare a recorded fallback. Submit at: https://openreview.net/group?id=AAAI.org/2027/Workshop/WMAC IMPORTANT DATES Paper submission deadline: November 20, 2026 (AoE) Notification of acceptance: December 2, 2026 Workshop: February 22–23, 2027, Montréal Dates follow the schedule set by the AAAI-27 workshop chairs and may change. Check the website for updates. PROGRAM A full day built for interaction: four invited talks, contributed oral and lightning talks, a poster session, breakout discussions in small groups, live industry demos, and a closing panel. Invited participants (tentative, pending confirmation): * Michael Wooldridge, University of Oxford * Samy Bengio, Apple / EPFL * Mikaela Grace, Anthropic * Bo An, Nanyang Technological University * Noah Goodman, Stanford University ORGANIZERS * Raphael Shu, OpenAgents (workshop chair) * Alborz Geramifard, LinkedIn * Alex Marin, Thomson Reuters Labs * Yusen Zhang, Columbia University * Asli Celikyilmaz, Microsoft / University of Washington ABOUT THE SERIES WMAC is the Workshop on Multi-Agent Collaboration at AAAI. WMAC 2025 (Philadelphia) received 33 submissions with 178 registered attendees; WMAC 2026 (Singapore, AAAI-26 Bridge Program) received 42 submissions with 360 registered attendees. Past editions: https://multiagents.org/ CONTACT Raphael Shu (raphael@uaca.com). We look forward to your submissions and to seeing you in Montréal! Best, ____________________________ Alborz Geramifard <https://alborz-geramifard.com/>