Public goods projects, such as open-source technology, are essential for the blockchain ecosystem's growth. However, funding these projects effectively remains a critical issue within the ecosystem. Currently, the funding protocols for blockchain public goods lack professionalism and fail to learn from past experiences. To address this challenge, our research introduces a human oracle protocol involving public goods projects, experts, and funders. In our approach, funders contribute investments to a funding pool, while experts offer investment advice based on their expertise in public goods projects. The oracle's decisions on funding support are influenced by the reputations of the experts. Experts earn or lose reputation based on how well their project implementations align with their advice, with successful investments leading to higher reputations. Our oracle is designed to adapt to changing circumstances, such as experts exiting or entering the decision-making process. We also introduce a regret bound to gauge the oracle's effectiveness. Theoretically, we establish an upper regret bound for both static and dynamic models and demonstrate its closeness to an asymptotically equal lower bound. Empirically, we implement our protocol on a test chain and show that our oracle's investment decisions closely mirror optimal investments in hindsight.

Knowledge transfer in cooperative multi-agent reinforcement learning (MARL) has drawn increasing attention in recent years. Unlike generalizing policies in single-agent tasks, it is more important to consider coordination knowledge than individual knowledge in multi-agent transfer learning. However, most of the existing methods only focus on knowledge transfer of the individual agent policy, which leads to coordination bias and finally affects the final performance in cooperative MARL. In this paper, we propose a level-adaptive MARL framework called “LA-QTransformer”, to realize the knowledge transfer on the coordination level via efficiently decomposing the agent coordination into multi-level coalition patterns for different agents. Compatible with centralized training with decentralized execution (CTDE) regime, LA-QTransformer utilizes the Level- Adaptive Transformer to generate suitable coalition patterns and then realizes the credit assignment for each agent. Besides, to deal with unexpected changes in the number of agents in the coordination transfer phase, we design a policy network called “Population invariant agent with Transformer (PIT)” to adapt dynamic observation and action space. We evaluate the LAQTransformer and PIT in the StarCraft II micro-management benchmark by comparing them with several state-of-the-art MARL baselines. The experimental results demonstrate the superiority of LA-QTransformer and PIT and verify the feasibility of coordination knowledge transfer.

We consider a hierarchical structure of a permissioned blockchain with three types of participant: providers, collectors, and governors. Providers forward transactions to collectors; collectors upload received transactions to governors after verifying and labeling them; and governors validate a portion of the labeled transactions they receive, pack valid transactions into a block, and append the block to the ledger. This model has various fields of application including data collection from the Internet-of-Things and second-hand markets. Our main contribution is to propose a reputation-based protocol to help governors evaluate the reliability of collectors. Specifically, given a transaction, each governor runs a softmax-based function to calculate a probability for each collector that sent and labeled this transaction. The probabilities, calculated using collectors’ reputations as inputs, represent the likelihood of the lead governor selecting the labeled transaction from collectors to consider for further validation. After the lead governor verifies a transaction, all collectors’ reputations are updated in line with the agreement of their labeling and the validity of the transaction as found by the lead governor. We show, both theoretically and empirically, that our protocol can significantly reduce governors’ verification workloads while maintaining firm liveness and high incentives.

Public goods projects, including open source technology, client development, and blockchain knowledge education, play an important role in the flourishing blockchain ecosystem. Accordingly, decision making for public goods funding is a key issue in the studies of the blockchain ecosystem. This work develops a human oracle protocol approach, involved with public goods projects, experts, and funders, as a solution to the public goods investment problem on blockchain. In our human oracle, funders contribute their investments, which are stored in a funding pool. Experts provide investment advice on public goods projects based on their experience. Decisions made by the human oracle on the amount of support from the funding pool are based on experts’ reputation. The reputation of each expert is updated by the performance of the project’s implementation in comparison to her advice. That is, better investment performance brings a higher reputation. Besides being applied to static model, our human oracle can also be extended to accommodate dynamic settings, in which the experts might leave or join the decision-making process. We introduce a regret bound to measure the effectiveness of our human oracle. Theoretically, we prove an upper regret bound for both static and dynamic models, and prove its tightness with an asymptotically equal lower bound. Empirically, we show that our oracle’s investment decision is close to the optimal investment in hindsight.

In many real-world cooperative multiagent reinforcement learning (MARL) tasks, teams of agents can rehearse together before deployment, but then communication constraints may force individual agents to execute independently when deployed. Centralized training and decentralized execution (CTDE) is increasingly popular in recent years, focusing mainly on this setting. In the value-based MARL branch, credit assignment mechanism is typically used to factorize the team reward into each individual’s reward — individual-global-max (IGM) is a condition on the factorization ensuring that agents’ action choices coincide with team’s optimal joint action. However, current architectures fail to consider local coordination within sub-teams that should be exploited for more effective factorization, leading to faster learning. We propose a novel value factorization framework, called multiagent Q-learning with sub-team coordination (QSCAN), to flexibly represent sub-team coordination while honoring the IGM condition. QSCAN encompasses the full spectrum of sub-team coordination according to sub-team size, ranging from the monotonic value function class to the entire IGM function class, with familiar methods such as QMIX and QPLEX located at the respective extremes of the spectrum. Experimental results show that QSCAN’s performance dominates state-of-the-art methods in matrix games, predator-prey tasks, the Switch challenge in MA-Gym. Additionally, QSCAN achieves comparable performances to those methods in a selection of StarCraft II micro-management tasks.

Permissioned blockchains take more reliability on participants than permissionless ones. In this poster, we focus on a hierarchical scenario of permissioned blockchains, which includes three types of participants: providers, collectors, and governors. Such a scenario has many applications in the field of IoT data collection, horizontal strategic alliances, etc. Our object is to reduce the cost of the governor's transaction verification. For this purpose, we propose a reputation protocol to help the governor measure the reliability of collectors. Based on the measurement of collectors' reputations, governors can pack high-quality transactions from reliable collectors into blocks, and thus the cost of verifying transactions can be decreased effectively. Through theoretical analysis, our protocol dramatically reduces the verification loss of governors.