P-FOLIO: Evaluating and Improving Logical Reasoning with Abundant Human-Written Reasoning Chains

Existing methods on understanding the capabilities of LLMs in logical reasoning rely on binary entailment classification or synthetically derived rationales, which are not sufficient for proper investigation of model's capabilities. We present P-FOLIO, a human-annotated dataset consisting of diverse and complex reasoning chains for a set of realistic logical reasoning stories also written by humans. P-FOLIO is collected with an annotation protocol that facilitates humans to annotate well-structured natural language proofs for first-order logic reasoning problems in a step-by-step manner. The number of reasoning steps in P-FOLIO span from 0 to 20. We further use P-FOLIO to evaluate and improve large-language-model (LLM) reasoning capabilities. We evaluate LLM reasoning capabilities at a fine granularity via single-step inference rule classification, with more diverse inference rules of more diverse and higher levels of complexities than previous works. Given that a single model-generated reasoning chain could take a completely different path than the human-annotated one, we sample multiple reasoning chains from a model and use pass@k metrics for evaluating the quality of model-generated reasoning chains. We show that human-written reasoning chains significantly boost the logical reasoning capabilities of LLMs via many-shot prompting and fine-tuning. Furthermore, fine-tuning Llama3-7B on P-FOLIO improves the model performance by 10% or more on three other out-of-domain logical reasoning datasets. We also conduct detailed analysis to show where most powerful LLMs fall short in reasoning. We will release the dataset and code publicly.

Blockchain-based Federated Recommendation with Incentive Mechanism

Nowadays, federated recommendation technology is rapidly evolving to help multiple organisations share data and train models while meeting user privacy, data security and government regulatory requirements. However, federated recommendation increases customer system costs such as power, computational and communication resources. Besides, federated recommendation systems are also susceptible to model attacks and data poisoning by participating malicious clients. Therefore, most customers are unwilling to participate in federated recommendation without any incentive. To address these problems, we propose a blockchain-based federated recommendation system with incentive mechanism to promote more trustworthy, secure, and efficient federated recommendation service. First, we construct a federated recommendation system based on NeuMF and FedAvg. Then we introduce a reverse auction mechanism to select optimal clients that can maximize the social surplus. Finally, we employ blockchain for on-chain evidence storage of models to ensure the safety of the federated recommendation system. The experimental results show that our proposed incentive mechanism can attract clients with superior training data to engage in the federal recommendation at a lower cost, which can increase the economic benefit of federal recommendation by 54.9\% while improve the recommendation performance. Thus our work provides theoretical and technological support for the construction of a harmonious and healthy ecological environment for the application of federal recommendation.

Solving Robotics Problems in Zero-Shot with Vision-Language Models

We introduce Wonderful Team, a multi-agent visual LLM (VLLM) framework for solving robotics problems in the zero-shot regime. By zero-shot we mean that, for a novel environment, we feed a VLLM an image of the robot's environment and a description of the task, and have the VLLM output the sequence of actions necessary for the robot to complete the task. Prior work on VLLMs in robotics has largely focused on settings where some part of the pipeline is fine-tuned, such as tuning an LLM on robot data or training a separate vision encoder for perception and action generation. Surprisingly, due to recent advances in the capabilities of VLLMs, this type of fine-tuning may no longer be necessary for many tasks. In this work, we show that with careful engineering, we can prompt a single off-the-shelf VLLM to handle all aspects of a robotics task, from high-level planning to low-level location-extraction and action-execution. Wonderful Team builds on recent advances in multi-agent LLMs to partition tasks across an agent hierarchy, making it self-corrective and able to effectively partition and solve even long-horizon tasks. Extensive experiments on VIMABench and real-world robotic environments demonstrate the system's capability to handle a variety of robotic tasks, including manipulation, visual goal-reaching, and visual reasoning, all in a zero-shot manner. These results underscore a key point: vision-language models have progressed rapidly in the past year, and should strongly be considered as a backbone for robotics problems going forward.

Cold Diffusion on the Replay Buffer: Learning to Plan from Known Good States

Learning from demonstrations (LfD) has successfully trained robots to exhibit remarkable generalization capabilities. However, many powerful imitation techniques do not prioritize the feasibility of the robot behaviors they generate. In this work, we explore the feasibility of plans produced by LfD. As in prior work, we employ a temporal diffusion model with fixed start and goal states to facilitate imitation through in-painting. Unlike previous studies, we apply cold diffusion to ensure the optimization process is directed through the agent's replay buffer of previously visited states. This routing approach increases the likelihood that the final trajectories will predominantly occupy the feasible region of the robot's state space. We test this method in simulated robotic environments with obstacles and observe a significant improvement in the agent's ability to avoid these obstacles during planning.

L2CEval: Evaluating Language-to-Code Generation Capabilities of Large Language Models

Recently, large language models (LLMs), especially those that are pretrained on code, have demonstrated strong capabilities in generating programs from natural language inputs in a few-shot or even zero-shot manner. Despite promising results, there is a notable lack of a comprehensive evaluation of these models language-to-code generation capabilities. Existing studies often focus on specific tasks, model architectures, or learning paradigms, leading to a fragmented understanding of the overall landscape. In this work, we present L2CEval, a systematic evaluation of the language-to-code generation capabilities of LLMs on 7 tasks across the domain spectrum of semantic parsing, math reasoning and Python programming, analyzing the factors that potentially affect their performance, such as model size, pretraining data, instruction tuning, and different prompting methods. In addition to assessing model performance, we measure confidence calibration for the models and conduct human evaluations of the output programs. This enables us to identify and analyze the typical failure modes across various tasks and models. L2CEval offers a comprehensive understanding of the capabilities and limitations of LLMs in language-to-code generation. We also release the evaluation framework and all model outputs, hoping to lay the groundwork for further future research in this domain.