Skip to yearly menu bar Skip to main content

Workshop Sat, Dec 12, 2026 • 6:00 PM – 3:00 AM AEDT Salle 241

Foundations of Agentic Systems Theory (FAST)

Erik Miehling ⋅ Madeline G. Reinecke ⋅ Hamza Mostafa ⋅ Jordan McAfoose ⋅ Irina Rish ⋅ Djallel Bouneffouf

[ OpenReview]

Abstract

The Foundations of Agentic Systems Theory (FAST) workshop provides a venue for the investigation of system-level behaviors of agentic AI. FAST aims to draw from various fields (including complex systems, developmental biology, organizational sociology, and cognitive science) with the goal of understanding which specific mechanisms of emergent behavior from other systems carry over to systems of LLM-based agents, which properties of the underlying agents (and their LLMs) facilitate or block these behaviors, and to what extent system-wide outcomes can be controlled or induced. Answering these questions is the prerequisite for safe and well-understood deployment of agentic AI.

Chat is not available.

Schedule

Time zone: Conference local time zone

Choosing LLMs for Strategic Agent Interactions

Arastun Mammadli ⋅ Herbert Woisetschläger ⋅ Jawad Fayaz ⋅ Shiqiang Wang

Choosing LLMs for Strategic Agent Interactions

Arastun Mammadli ⋅ Herbert Woisetschläger ⋅ Jawad Fayaz ⋅ Shiqiang Wang

Group size effects and collective misalignment in LLM multi-agent systems

Ariel Flint ⋅ Luca Maria Aiello ⋅ Romualdo Pastor-Satorras ⋅ Andrea Baronchelli

Group size effects and collective misalignment in LLM multi-agent systems

Ariel Flint ⋅ Luca Maria Aiello ⋅ Romualdo Pastor-Satorras ⋅ Andrea Baronchelli

Cooperative Mechanism Games for Oversight of Agent Populations under Partitioned Observation

Joel Christoph

Cooperative Mechanism Games for Oversight of Agent Populations under Partitioned Observation

Joel Christoph

Building Autonomous Causal Analysis Agents from Narrative Data

Hongxin Zhen ⋅ Philip Torr ⋅ Jialin Yu

Building Autonomous Causal Analysis Agents from Narrative Data

Hongxin Zhen ⋅ Philip Torr ⋅ Jialin Yu

Moral Hazard in Multi-Agent Language Models

Dane Malenfant

Moral Hazard in Multi-Agent Language Models

Dane Malenfant

What Compression Preserves: Instruction Survival in Long-Horizon LLM Agents

Daniel Kirste ⋅ Alexander Kropiunig

What Compression Preserves: Instruction Survival in Long-Horizon LLM Agents

Daniel Kirste ⋅ Alexander Kropiunig

When Equivalent Representations Produce Different Responses: Prompt-Sensitive Stackelberg Games with LLM Followers

Xander Barron ⋅ Weici Pan ⋅ Jiawei Zhou ⋅ Zhenhua Liu

When Equivalent Representations Produce Different Responses: Prompt-Sensitive Stackelberg Games with LLM Followers

Xander Barron ⋅ Weici Pan ⋅ Jiawei Zhou ⋅ Zhenhua Liu

When Does Composition Preserve Coherence? A Product-Structure Criterion and a Magnitude Law for Multi-Agent Belief Assembly

Anany Kotawala

When Does Composition Preserve Coherence? A Product-Structure Criterion and a Magnitude Law for Multi-Agent Belief Assembly

Anany Kotawala

Toward collective intelligence: evolutionary pressures for cooperative language models

Mariana Meireles ⋅ Terry J Zhang ⋅ David Guzman Piedrahita ⋅ Zhijing Jin

Toward collective intelligence: evolutionary pressures for cooperative language models

Mariana Meireles ⋅ Terry J Zhang ⋅ David Guzman Piedrahita ⋅ Zhijing Jin

Safety Composes, but Its Proof Does Not: Exponential Proof-Observability Gaps for Agent Contracts

Manoj Saravanan

Safety Composes, but Its Proof Does Not: Exponential Proof-Observability Gaps for Agent Contracts

Manoj Saravanan

A Taxonomy of Agentic Errors: A systematic review of how agentic AI systems can fail

David Piorkowski ⋅ Michelle Brachman ⋅ Naim Zierau ⋅ Adriana A Garcia ⋅ Jessica He

A Taxonomy of Agentic Errors: A systematic review of how agentic AI systems can fail

David Piorkowski ⋅ Michelle Brachman ⋅ Naim Zierau ⋅ Adriana A Garcia ⋅ Jessica He

Analyzing The Changes of AI Safety Language Across 4 Frontier Labs

Yejin Kim ⋅ Rachita Mehta ⋅ Sakshi Chaubey ⋅ Lisa Villeneuve ⋅ Caitlin L Mah

Analyzing The Changes of AI Safety Language Across 4 Frontier Labs

Yejin Kim ⋅ Rachita Mehta ⋅ Sakshi Chaubey ⋅ Lisa Villeneuve ⋅ Caitlin L Mah

Instrumental Convergence Revisited: Towards a Broader Characterization of Runaway Optimization

Mateusz Bagiński ⋅ Tushant Jha

Instrumental Convergence Revisited: Towards a Broader Characterization of Runaway Optimization

Mateusz Bagiński ⋅ Tushant Jha

Individually Sensible, Collectively Harmful: LLM Agents in Repeated Commons Games

Yujiao Chen

Individually Sensible, Collectively Harmful: LLM Agents in Repeated Commons Games

Yujiao Chen

Black-Box Auditing of Hidden Loyalties in Agent-Bound Checkpoints with Matched Entity Swaps

Martin Ciesielski-Listwan ⋅ Alyssia Jovellanos

Black-Box Auditing of Hidden Loyalties in Agent-Bound Checkpoints with Matched Entity Swaps

Martin Ciesielski-Listwan ⋅ Alyssia Jovellanos

Alignment and Candor in AI Delegation

Mikayel Sukiasyan

Alignment and Candor in AI Delegation

Mikayel Sukiasyan

Stress-Testing Context Alignment Stressors in Agents

Mark Lovett ⋅ Syed A Haider ⋅ Christopher Denq ⋅ Promita R Sikder ⋅ Elizaveta Tennant

Stress-Testing Context Alignment Stressors in Agents

Mark Lovett ⋅ Syed A Haider ⋅ Christopher Denq ⋅ Promita R Sikder ⋅ Elizaveta Tennant

Reciprocal Wireheading Undermines Alignment Properties in Multi-Agent LLMs Evaluations

Vayk Mathrani ⋅ Saif Ur Rehman ⋅ Ahmad Shaker ⋅ Thao Pham

Reciprocal Wireheading Undermines Alignment Properties in Multi-Agent LLMs Evaluations

Vayk Mathrani ⋅ Saif Ur Rehman ⋅ Ahmad Shaker ⋅ Thao Pham

What Makes an LLM a Good Optimizer? A Trajectory Analysis of LLM-Guided Evolutionary Search

Xinhao Zhang ⋅ Xi Chen ⋅ François Portet ⋅ Maxime Peyrard

What Makes an LLM a Good Optimizer? A Trajectory Analysis of LLM-Guided Evolutionary Search

Xinhao Zhang ⋅ Xi Chen ⋅ François Portet ⋅ Maxime Peyrard

Cooperative Words, Selfish Actions: How Strategic Optimization Shapes LLMs' Interaction With Humans

Yaomin Jiang ⋅ Yannik Keller ⋅ Levin Brinkmann ⋅ Tobias Werner ⋅ Seyed Ali Alhosseini ⋅ Michiel Bakker ⋅ Jean-François Bonnefon ⋅ Iyad Rahwan

Cooperative Words, Selfish Actions: How Strategic Optimization Shapes LLMs' Interaction With Humans

Yaomin Jiang ⋅ Yannik Keller ⋅ Levin Brinkmann ⋅ Tobias Werner ⋅ Seyed Ali Alhosseini ⋅ Michiel Bakker ⋅ Jean-François Bonnefon ⋅ Iyad Rahwan

When Behavioral Signal Is Not Enough: Separating Observability from Learned Theory-of-Mind Inference

Yunus Emre Bulut

When Behavioral Signal Is Not Enough: Separating Observability from Learned Theory-of-Mind Inference

Yunus Emre Bulut

Nudging the Graph, Not the Agent: A Benchmark for Alignment by Environment Design

Yuriy Dementiev

Nudging the Graph, Not the Agent: A Benchmark for Alignment by Environment Design

Yuriy Dementiev

The Weight of Teamwork: Does Communication Structure Help LLM Scientists Search for Scientific Laws?

Jianing Shi ⋅ Zaibin Zhang ⋅ Shijun Lei ⋅ Zhenfei Yin

The Weight of Teamwork: Does Communication Structure Help LLM Scientists Search for Scientific Laws?

Jianing Shi ⋅ Zaibin Zhang ⋅ Shijun Lei ⋅ Zhenfei Yin

Coherence-Driven Belief Formation and Population Dynamics of Contagion in LLM Agents

Tathagata Banerjee ⋅ Nima Moghaddas

Coherence-Driven Belief Formation and Population Dynamics of Contagion in LLM Agents

Tathagata Banerjee ⋅ Nima Moghaddas

Collusion as a Causal Estimand

Junlin Chen

Collusion as a Causal Estimand

Junlin Chen

Modal Agency: Constraints, Affordances, and Self-Constructing Harnesses

Christian Gensbigler

Modal Agency: Constraints, Affordances, and Self-Constructing Harnesses

Christian Gensbigler

Before Agents Decide: Epistemic Action in LLM-Based Systems

YIZHI LIU ⋅ Balaji Padmanabhan ⋅ Siva Viswanathan

Before Agents Decide: Epistemic Action in LLM-Based Systems

YIZHI LIU ⋅ Balaji Padmanabhan ⋅ Siva Viswanathan

Replaying Self-Evolution's Tape: Understanding Self-Evolving Agents Requires Experimental Evolution

Xinhao Zhang ⋅ Kristina Gligoric ⋅ François Portet ⋅ Maxime Peyrard

Replaying Self-Evolution's Tape: Understanding Self-Evolving Agents Requires Experimental Evolution

Xinhao Zhang ⋅ Kristina Gligoric ⋅ François Portet ⋅ Maxime Peyrard

Reference Misgeneralization: Why Orthogonality Thesis Doesn’t Go Far Enough, and the Problem of Deep Brittleness

Jiske de Vries ⋅ Sahil Kulshrestha ⋅ Ana Simina Stoian ⋅ Tushant Jha

Reference Misgeneralization: Why Orthogonality Thesis Doesn’t Go Far Enough, and the Problem of Deep Brittleness

Jiske de Vries ⋅ Sahil Kulshrestha ⋅ Ana Simina Stoian ⋅ Tushant Jha

Model Heterogeneity Attenuates Escalation Dynamics among Large Language Models in War Games

Jihyeon Hyeong ⋅ Hiromu Yakura ⋅ Iyad Rahwan

Model Heterogeneity Attenuates Escalation Dynamics among Large Language Models in War Games

Jihyeon Hyeong ⋅ Hiromu Yakura ⋅ Iyad Rahwan

Knowing When to Speak: State-Aware Participation in Multi-Agent LLM Systems

Xingze Wang

Knowing When to Speak: State-Aware Participation in Multi-Agent LLM Systems

Xingze Wang

Kuramoto Synchronization in LLM-Based Multi-Agent Systems

Francesco Bertolotti ⋅ Francesco Nappo

Kuramoto Synchronization in LLM-Based Multi-Agent Systems

Francesco Bertolotti ⋅ Francesco Nappo

Is the Agent Aware of the Situation? Probing Embodied DRL Agents' Situation Assessment

Elahe Soltandoost ⋅ Andrea Salfinger ⋅ Lauro Snidaro

Is the Agent Aware of the Situation? Probing Embodied DRL Agents' Situation Assessment

Elahe Soltandoost ⋅ Andrea Salfinger ⋅ Lauro Snidaro

When Does Group Structure Emerge? Validating History-Dependent Group Behavior in LLM Agent Populations

Yujiao Chen ⋅ Jean Farmer

When Does Group Structure Emerge? Validating History-Dependent Group Behavior in LLM Agent Populations

Yujiao Chen ⋅ Jean Farmer

Harness as a Language: A Minimalist Agent Framework With Maximal Expressivity

Zhening Li ⋅ Joshua Liu ⋅ Mateja Vukelic ⋅ Nicole Shen ⋅ Supriya Lall ⋅ Amitayush Thakur ⋅ Alex Zhang ⋅ Omar Khattab ⋅ Jonathan Li ⋅ Armando Solar-Lezama

Harness as a Language: A Minimalist Agent Framework With Maximal Expressivity

Zhening Li ⋅ Joshua Liu ⋅ Mateja Vukelic ⋅ Nicole Shen ⋅ Supriya Lall ⋅ Amitayush Thakur ⋅ Alex Zhang ⋅ Omar Khattab ⋅ Jonathan Li ⋅ Armando Solar-Lezama

Nothing Is Harder Than Nothing, Until You Write It Down: Absence Conventions in Stateful Tool Use

Marek Suppa ⋅ Jaroslav Kopcan

Nothing Is Harder Than Nothing, Until You Write It Down: Absence Conventions in Stateful Tool Use

Marek Suppa ⋅ Jaroslav Kopcan

Critical Dynamics of AI Self-Improvement

Mikhail Burtsev

Critical Dynamics of AI Self-Improvement

Mikhail Burtsev

Logographic Emergent Communication In A Multi-Agent Sketching Lewis Game

Casper Dancy ⋅ Daniela Mihai ⋅ Jonathon Hare

Logographic Emergent Communication In A Multi-Agent Sketching Lewis Game

Casper Dancy ⋅ Daniela Mihai ⋅ Jonathon Hare

The Question of Where: A Placement Asymmetry in How Agents Learn to Cooperate

Samridh Aggarwal

The Question of Where: A Placement Asymmetry in How Agents Learn to Cooperate

Samridh Aggarwal

Group selection to evolve cooperative prompts in populations of LLM agents

Luis Celiktemel ⋅ Edward Eichhorn ⋅ Levin Brinkmann ⋅ Robin Schimmelpfennig ⋅ Aron Vallinder ⋅ Yaomin Jiang ⋅ Edward Hughes ⋅ Iyad Rahwan

Group selection to evolve cooperative prompts in populations of LLM agents

Luis Celiktemel ⋅ Edward Eichhorn ⋅ Levin Brinkmann ⋅ Robin Schimmelpfennig ⋅ Aron Vallinder ⋅ Yaomin Jiang ⋅ Edward Hughes ⋅ Iyad Rahwan

Placement, Not Variance: Redefining the Axis of Security for Multi-Agent AI Systems

Zakaria Elghazzali ⋅ Kenneth Eaton

Placement, Not Variance: Redefining the Axis of Security for Multi-Agent AI Systems

Zakaria Elghazzali ⋅ Kenneth Eaton

Deal With It! Diagnosing Post-Training in Bargaining, from Fluency to Surplus Extraction

Mohid Butt

Deal With It! Diagnosing Post-Training in Bargaining, from Fluency to Surplus Extraction

Mohid Butt

Code Comments as an Alignment Surface: Stigmergic Steering of Coding Agents

Jonathan Eicher ⋅ Rafael F Irgolič

Code Comments as an Alignment Surface: Stigmergic Steering of Coding Agents

Jonathan Eicher ⋅ Rafael F Irgolič

Aggregating Open-Ended Answers Across Populations of LLM Agents

Sneheel Sarangi ⋅ Kevin Chou ⋅ Haocheng Guo ⋅ Ahan Karnik ⋅ Vihaan Kabra

Aggregating Open-Ended Answers Across Populations of LLM Agents

Sneheel Sarangi ⋅ Kevin Chou ⋅ Haocheng Guo ⋅ Ahan Karnik ⋅ Vihaan Kabra

What Is a Morally Aligned AI Agent? A Philosophical Coverage Map and Its Failure Under Composition

Syed A Haider

What Is a Morally Aligned AI Agent? A Philosophical Coverage Map and Its Failure Under Composition

Syed A Haider

Who Should Act After Silence? Exact Coordination Policies for Agents on Unreliable Relay Lines

Andrei Ermakov

Who Should Act After Silence? Exact Coordination Policies for Agents on Unreliable Relay Lines

Andrei Ermakov

Norms at a Price: Why RL-Based Alignment Can Promise Conditional Compliance at Best

Kevin Baum ⋅ Ruta Binkyte ⋅ Felix Jahn

Norms at a Price: Why RL-Based Alignment Can Promise Conditional Compliance at Best

Kevin Baum ⋅ Ruta Binkyte ⋅ Felix Jahn

Subagents vs Agent Skills: Executing Reusable Knowledge for Long-Horizon Agentic Tasks

Top Piriyakulkij ⋅ Rachel Lawrence ⋅ Alicia Curth ⋅ Sushrut Karmalkar ⋅ Niranjani Prasad

Subagents vs Agent Skills: Executing Reusable Knowledge for Long-Horizon Agentic Tasks

Top Piriyakulkij ⋅ Rachel Lawrence ⋅ Alicia Curth ⋅ Sushrut Karmalkar ⋅ Niranjani Prasad

What Counts as the Agent? Boundaries and Measurement in Agentic Systems

Sergio Alvarez-Napagao ⋅ Lara Ahmed ⋅ Luis Oliva-Felipe ⋅ Júlia López-Closa ⋅ Sara Montese ⋅ Ulises Cortés ⋅ Victor Gimenez-Abalos

What Counts as the Agent? Boundaries and Measurement in Agentic Systems

Sergio Alvarez-Napagao ⋅ Lara Ahmed ⋅ Luis Oliva-Felipe ⋅ Júlia López-Closa ⋅ Sara Montese ⋅ Ulises Cortés ⋅ Victor Gimenez-Abalos

Sequential Social Inference in Large Language Models

Elif Akata ⋅ Dylan Cope ⋅ Eric Schulz ⋅ Jakob Foerster

Sequential Social Inference in Large Language Models

Elif Akata ⋅ Dylan Cope ⋅ Eric Schulz ⋅ Jakob Foerster

Beyond Accuracy: Formalization for Understanding Information Exchange in Multi-Agent Debate

Sarah Wu ⋅ Tycho Young

Beyond Accuracy: Formalization for Understanding Information Exchange in Multi-Agent Debate

Sarah Wu ⋅ Tycho Young

The Observability Gap in Multi-Agent AI Systems

Eric Mibuari ⋅ Madhu Srikumar ⋅ Claire R Leibowicz ⋅ Sarah Tan ⋅ Borhane Blili-Hamelin ⋅ Vinh X Nguyen ⋅ Amin Oueslati ⋅ Xing Han Lu ⋅ Jason Stanley ⋅ Sean McGregor ⋅ Kyle W Hall ⋅ Kevin Klyman ⋅ Abhi Sanka

The Observability Gap in Multi-Agent AI Systems

Eric Mibuari ⋅ Madhu Srikumar ⋅ Claire R Leibowicz ⋅ Sarah Tan ⋅ Borhane Blili-Hamelin ⋅ Vinh X Nguyen ⋅ Amin Oueslati ⋅ Xing Han Lu ⋅ Jason Stanley ⋅ Sean McGregor ⋅ Kyle W Hall ⋅ Kevin Klyman ⋅ Abhi Sanka

Rational Clarification by Assistive Agents via Value-of-Information Reasoning

T. Duy Nguyen-Hien ⋅ Yee Whye Teh ⋅ Wee Sun Lee ⋅ Tan Zhi-Xuan

Rational Clarification by Assistive Agents via Value-of-Information Reasoning

T. Duy Nguyen-Hien ⋅ Yee Whye Teh ⋅ Wee Sun Lee ⋅ Tan Zhi-Xuan

In & Out: Social Dilemmas With Frames and Self-Negotiated Contracts

Xuanqiang A Huang ⋅ Charlie Tharas ⋅ Samuele Marro ⋅ Van Q. Truong ⋅ Bernhard Schölkopf ⋅ Emanuele La Malfa ⋅ Zhijing Jin

In & Out: Social Dilemmas With Frames and Self-Negotiated Contracts

Xuanqiang A Huang ⋅ Charlie Tharas ⋅ Samuele Marro ⋅ Van Q. Truong ⋅ Bernhard Schölkopf ⋅ Emanuele La Malfa ⋅ Zhijing Jin

Protoreasoning in Tiny Transformers

Eduardo Valle ⋅ Fergal Reid

Protoreasoning in Tiny Transformers

Eduardo Valle ⋅ Fergal Reid

Moral Surface Without Moral Character: Value Labels Do Not Determine LLM Agent Behavior

Syed A Haider

Moral Surface Without Moral Character: Value Labels Do Not Determine LLM Agent Behavior

Syed A Haider

Belief Attribution Requires Behavioral Evidence and Agents Now Supply It

Syed A Haider

Belief Attribution Requires Behavioral Evidence and Agents Now Supply It

Syed A Haider

Beyond Machina Economicus: On Cognitive Organization and Efficiency Arguments

Alessandro Veneri ⋅ Kuil Schoneveld ⋅ Ka Wang Pang ⋅ Tushant Jha

Beyond Machina Economicus: On Cognitive Organization and Efficiency Arguments

Alessandro Veneri ⋅ Kuil Schoneveld ⋅ Ka Wang Pang ⋅ Tushant Jha