LLM Routing and Aggregation

Leveraging multiple LLMs to improve collective performance and efficiency.

1. Beyond Gemini-3-Pro: Revisiting LLM Routing and Aggregation at Scale (arXiv)
2. LLMRouterBench: A Massive Benchmark and Unified Framework for LLM Routing (arXiv)
3. Beyond GPT-5: Making LLMs Cheaper and Better via Performance-Efficiency Optimized Routing (paper)
4. ICL-Router: In-Context Learned Model Representations for LLM Routing (arXiv)
5. Do We Truly Need So Many Samples? Multi-LLM Repeated Sampling Efficiently Scales Test-Time Compute (arXiv)
6. The Avengers: A Simple Recipe for Uniting Smaller Language Models to Challenge Proprietary Giants (arXiv)
7. Open-Source LLMs Collaboration Beats Closed-Source LLMs: A Scalable Multi-Agent System (arXiv)
8. Learning Compact Representations of LLM Abilities via Item Response Theory (arXiv)
9. Decouple and Orthogonalize: A Data-Free Framework for LoRA Merging (arXiv)
10. Nature-Inspired Population-Based Evolution of Large Language Models (arXiv)
11. A Comprehensive Survey of LLM-Driven Collective Intelligence: Past, Present, and Future (paper)

LLM-based Agent

Agent architectures, memory systems, workflow generation, and long-horizon task execution for LLM-based agents.

1. Single-Agent Scaling Fails Multi-Agent Intelligence: Towards Foundation Models with Native Multi-Agent Intelligence (arXiv)
2. If Multi-Agent Debate is the Answer, What is the Question? (arXiv)
3. MemVerse: Multimodal Memory for Lifelong Learning Agents (arXiv)
4. EvoFlow: Evolving Diverse Agentic Workflows On The Fly (arXiv)
5. PerPilot: Personalizing VLM-based Mobile Agents via Memory and Exploration (arXiv)
6. Organizing, Orchestrating, and Benchmarking Agent Skills at Ecosystem Scale (arXiv)
7. InternAgent-1.5: A Unified Agentic Framework for Long-Horizon Autonomous Scientific Discovery (arXiv)
8. Hands-on LLM-based Agents: A Tutorial for General Audiences (paper)

LLM-based MAS

Multi-agent mechanisms built with LLM-based agents, focusing on norms, reputation, and social interaction.

1. Beyond the Tragedy of the Commons: Building A Reputation System for Generative Multi-agent Systems (arXiv)
2. Emergence of Social Norms in Generative Agent Societies: Principles and Architecture (arXiv)
3. OASIS: Open Agent Social Interaction Simulations with One Million Agents (arXiv)
4. An AI Researchers' Perspective: At the Crossroad of LLMs, Agent-Based Modeling, and Complex Systems (paper)

LLM Reasoning and Reinforcement Learning

Methods that improve LLM reasoning, mostly through the lens of reinforcement learning.

1. Reinforcement Learning for Large Language Models via Group Preference Reward Shaping (paper)
2. Squeeze the Soaked Sponge: Efficient Off-policy Reinforcement Finetuning for Large Language Model (arXiv)
3. The Path of Self-Evolving Large Language Models: Achieving Data-Efficient Learning via Intrinsic Feedback (arXiv)
4. Scaling Physical Reasoning with the PHYSICS Dataset (arXiv)
5. ReMA: Learning to Meta-think for LLMs with Multi-Agent Reinforcement Learning (arXiv)
6. PhysicsMinions: Winning Gold Medals in the Latest Physics Olympiads with a Coevolutionary Multimodal Multi-Agent System (arXiv)
7. MARTI-MARS: Scaling Multi-Agent Self-Search via Reinforcement Learning for Code Generation (arXiv)

Multi-Agent Learning in Games

Theoretical analyses of multi-agent learning dynamics and equilibrium behavior.

1. A Formal Model for Multiagent Q-Learning on Graphs (paper)
2. Provably Efficient Information-Directed Sampling Algorithms for Multi-Agent Reinforcement Learning (paper)
3. Payoff Control in Multichannel Games: Influencing Opponent Learning Evolution (paper)
4. Regret Minimization in Population Network Games: Vanishing Heterogeneity and Convergence to Equilibria (paper)
5. The Stochastic Evolutionary Dynamics of Softmax Policy Gradient in Games (paper)
6. A Pair-Approximation Method for Modelling the Dynamics of Multi-Agent Stochastic Games (record)
7. Learning by Reusing Previous Advice: A Memory-Based Teacher-Student Framework (paper)
8. The Best of Both Worlds in Network Population Games: Reaching Consensus and Convergence to Equilibrium (paper)
9. Individual-Level Inverse Reinforcement Learning for Mean Field Games (paper)
10. Modelling the Dynamics of Multi-Agent Q-Learning: The Stochastic Effects of Local Interaction and Incomplete Information (paper)
11. Modelling the Dynamics of Regret Minimization in Large Agent Populations: A Master Equation Approach (paper)
12. The Dynamics of Q-Learning in Population Games: A Physics-Inspired Continuity Equation Model (paper)
13. Dynamics of Q-Learning in Networked Stochastic Games (paper)

Cooperation in Social Dilemmas

Understanding cooperation emergence in social dilemma games.

1. A Successful Strategy for Multichannel Iterated Prisoner's Dilemma (paper)
2. Multi-Agent, Human-Agent and Beyond: A Survey on Cooperation in Social Dilemmas (paper)
3. Emergence of Punishment in Social Dilemma with Environmental Feedback (record)
4. Facilitating Cooperation in Human-Agent Hybrid Populations through Autonomous Agents (paper)
5. How Committed Individuals Shape Social Dynamics: A Survey on Coordination Games and Social Dilemma Games (paper)
6. Beyond a Binary Theorizing of Prosociality (paper)
7. Large Language Models Overcome the Machine Penalty When Acting Fairly but Not When Acting Selfishly or Altruistically (arXiv)

LLM Meets Psychology

1. Large Language Models are Near-Optimal Decision-Makers with a Non-Human Learning Behavior (arXiv)