arXiv subjects

Xujie Si

Publications and source records attributed to Xujie Si.

At least 37 records · Page 2Linked to original sources

arXiv2025

Understanding Behavioral Metric Learning: A Large-Scale Study on Distracting Reinforcement Learning Environments

A key approach to state abstraction is approximating behavioral metrics (notably, bisimulation metrics) in the observation space and embedding these learned distances in the representation space. While promising for robustness to task-irrelevant noise, as shown in prior work, accurately estimating these metrics remains challenging, requiring various design choices that create gaps between theory and practice. Prior evaluations focus mainly on final returns, leaving the quality of learned metrics and the source of performance gains unclear. To systematically assess how metric learning works in deep reinforcement learning (RL), we evaluate five recent approaches, unified conceptually as isometric embeddings with varying design choices. We benchmark them with baselines across 20 state-based and 14 pixel-based tasks, spanning 370 task configurations with diverse noise settings. Beyond final returns, we introduce the evaluation of a denoising factor to quantify the encoder's ability to filter distractions. To further isolate the effect of metric learning, we propose and evaluate an isolated metric estimation setting, in which the encoder is influenced solely by the metric loss. Finally, we release an open-source, modular codebase to improve reproducibility and support future research on metric learning in deep RL.

cs.LG↗

arXiv2025

Towards Repository-Level Program Verification with Large Language Models

Recent advancements in large language models (LLMs) suggest great promises in code and proof generations. However, scaling automated formal verification to real-world projects requires resolving cross-module dependencies and global contexts, which are crucial challenges overlooked by existing LLM-based methods with a special focus on targeting isolated, function-level verification tasks. To systematically explore and address the significant challenges of verifying entire software repositories, we introduce RVBench, the first verification benchmark explicitly designed for repository-level evaluation, constructed from four diverse and complex open-source Verus projects. We further introduce RagVerus, an extensible framework that synergizes retrieval-augmented generation with context-aware prompting to automate proof synthesis for multi-module repositories. RagVerus triples proof pass rates on existing benchmarks under constrained model inference budgets, and achieves a 27% relative improvement on the more challenging RVBench benchmark, demonstrating a scalable and sample-efficient verification solution.

cs.SE↗

arXiv2025

A Case Study on the Effectiveness of LLMs in Verification with Proof Assistants

Large language models (LLMs) can potentially help with verification using proof assistants by automating proofs. However, it is unclear how effective LLMs are in this task. In this paper, we perform a case study based on two mature Rocq projects: the hs-to-coq tool and Verdi. We evaluate the effectiveness of LLMs in generating proofs by both quantitative and qualitative analysis. Our study finds that: (1) external dependencies and context in the same source file can significantly help proof generation; (2) LLMs perform great on small proofs but can also generate large proofs; (3) LLMs perform differently on different verification projects; and (4) LLMs can generate concise and smart proofs, apply classical techniques to new definitions, but can also make odd mistakes.

cs.PL↗

arXiv2025

TypyBench: Evaluating LLM Type Inference for Untyped Python Repositories

Type inference for dynamic languages like Python is a persistent challenge in software engineering. While large language models (LLMs) have shown promise in code understanding, their type inference capabilities remain underexplored. We introduce TypyBench, a benchmark designed to evaluate LLMs' type inference across entire Python repositories. TypyBench features two novel metrics: TypeSim, which captures nuanced semantic relationships between predicted and ground truth types, and TypeCheck, which assesses type consistency across codebases. Our evaluation of various LLMs on a curated dataset of 50 high-quality Python repositories reveals that, although LLMs achieve decent TypeSim scores, they struggle with complex nested types and exhibit significant type consistency errors. These findings suggest that future research should shift focus from improving type similarity to addressing repository-level consistency. TypyBench provides a foundation for this new direction, offering insights into model performance across different type complexities and usage contexts. Our code and data are available at https://github.com/typybench/typybench.

cs.SE↗

arXiv2025

$τ^2$-Bench: Evaluating Conversational Agents in a Dual-Control Environment

Existing benchmarks for conversational AI agents simulate single-control environments, where only the AI agent can use tools to interact with the world, while the user remains a passive information provider. This differs from real-world scenarios like technical support, where users need to actively participate in modifying the state of the (shared) world. In order to address this gap, we introduce $τ^2$-bench, with four key contributions: 1) A novel Telecom dual-control domain modeled as a Dec-POMDP, where both agent and user make use of tools to act in a shared, dynamic environment that tests both agent coordination and communication, 2) A compositional task generator that programmatically creates diverse, verifiable tasks from atomic components, ensuring domain coverage and controlled complexity, 3) A reliable user simulator tightly coupled with the environment, whose behavior is constrained by tools and observable states, improving simulation fidelity, 4) Fine-grained analysis of agent performance through multiple ablations including separating errors arising from reasoning vs communication/coordination. In particular, our experiments show significant performance drops when agents shift from no-user to dual-control, highlighting the challenges of guiding users. Overall, $τ^2$-bench provides a controlled testbed for agents that must both reason effectively and guide user actions.

cs.AI↗

arXiv2025

LogiCity: Advancing Neuro-Symbolic AI with Abstract Urban Simulation

Recent years have witnessed the rapid development of Neuro-Symbolic (NeSy) AI systems, which integrate symbolic reasoning into deep neural networks. However, most of the existing benchmarks for NeSy AI fail to provide long-horizon reasoning tasks with complex multi-agent interactions. Furthermore, they are usually constrained by fixed and simplistic logical rules over limited entities, making them far from real-world complexities. To address these crucial gaps, we introduce LogiCity, the first simulator based on customizable first-order logic (FOL) for an urban-like environment with multiple dynamic agents. LogiCity models diverse urban elements using semantic and spatial concepts, such as IsAmbulance(X) and IsClose(X, Y). These concepts are used to define FOL rules that govern the behavior of various agents. Since the concepts and rules are abstractions, they can be universally applied to cities with any agent compositions, facilitating the instantiation of diverse scenarios. Besides, a key feature of LogiCity is its support for user-configurable abstractions, enabling customizable simulation complexities for logical reasoning. To explore various aspects of NeSy AI, LogiCity introduces two tasks, one features long-horizon sequential decision-making, and the other focuses on one-step visual reasoning, varying in difficulty and agent behaviors. Our extensive evaluation reveals the advantage of NeSy frameworks in abstract reasoning. Moreover, we highlight the significant challenges of handling more complex abstractions in long-horizon multi-agent scenarios or under high-dimensional, imbalanced data. With its flexible design, various features, and newly raised challenges, we believe LogiCity represents a pivotal step forward in advancing the next generation of NeSy AI. All the code and data are open-sourced at our website: https://jaraxxus-me.github.io/LogiCity/

cs.AI↗

arXiv2025

LLM Library Learning Fails: A LEGO-Prover Case Study

Recent advancements in the coding, reasoning, and tool-using abilities of LLMs have spurred interest in library learning (i.e., online learning through the creation, storage, and retrieval of reusable and composable functions, knowledge, checklists, or lemmas). Such systems often promise improved task performance through the automatic creation of broadly applicable tools, as well as superior computational performance through the caching of reasoning (i.e., the storage of generated tools). However, we find strong reason to be skeptical. We perform a deep dive into one such system, LEGO-Prover, which purports to learn reusable lemmas for mathematical reasoning. We find no evidence of the direct reuse of learned lemmas, and find evidence against the soft reuse of learned lemmas (i.e., reuse by modifying relevant examples). Crucially, we find that LEGO-Prover does not in fact improve over the simple baseline of prompting the model - the improvements in task accuracy vanish once computational cost is accounted for. Our findings suggest that serious misconceptions exist as to the effectiveness of these techniques, that a serious re-examination of the state of LLM-based library learning is required, and that we require much stronger standards for evaluation including behavioural analysis and ensuring that an equal computational budget is used for baselines.

cs.LG↗

arXiv2025

Learning Minimal Neural Specifications

Formal verification is only as good as the specification of a system, which is also true for neural network verification. Existing specifications follow the paradigm of data as specification, where the local neighborhood around a reference data point is considered correct or robust. While these specifications provide a fair testbed for assessing model robustness, they are too restrictive for verifying any unseen test data points, a challenging task with significant real-world implications. Recent work shows great promise through a new paradigm, neural representation as specification, which uses neural activation patterns (NAPs) for this purpose. However, it computes the most refined NAPs, which include many redundant neurons. In this paper, we study the following problem: Given a neural network, find a minimal (general) NAP specification that is sufficient for formal verification of its robustness properties. Finding the minimal NAP specification not only expands verifiable bounds but also provides insights into which set of neurons contributes to the model's robustness. To address this problem, we propose three approaches: conservative, statistical, and optimistic. Each of these methods offers distinct strengths and trade-offs in terms of minimality and computational speed, making them suitable for scenarios with different priorities. Notably, the optimistic approach can probe potential causal links between neurons and the robustness of large vision neural networks without relying on verification tools, a task existing methods struggle to scale. Our experiments show that minimal NAP specifications use far fewer neurons than those from previous work while expanding verifiable boundaries by several orders of magnitude.

cs.LG↗

arXiv2025

Proving Olympiad Inequalities by Synergizing LLMs and Symbolic Reasoning

Large language models (LLMs) can prove mathematical theorems formally by generating proof steps (\textit{a.k.a.} tactics) within a proof system. However, the space of possible tactics is vast and complex, while the available training data for formal proofs is limited, posing a significant challenge to LLM-based tactic generation. To address this, we introduce a neuro-symbolic tactic generator that synergizes the mathematical intuition learned by LLMs with domain-specific insights encoded by symbolic methods. The key aspect of this integration is identifying which parts of mathematical reasoning are best suited to LLMs and which to symbolic methods. While the high-level idea of neuro-symbolic integration is broadly applicable to various mathematical problems, in this paper, we focus specifically on Olympiad inequalities (Figure~1). We analyze how humans solve these problems and distill the techniques into two types of tactics: (1) scaling, handled by symbolic methods, and (2) rewriting, handled by LLMs. In addition, we combine symbolic tools with LLMs to prune and rank the proof goals for efficient proof search. We evaluate our framework on 161 challenging inequalities from multiple mathematics competitions, achieving state-of-the-art performance and significantly outperforming existing LLM and symbolic approaches without requiring additional training data.

cs.AI↗

arXiv2025

RAG-Verus: Repository-Level Program Verification with LLMs using Retrieval Augmented Generation

Scaling automated formal verification to real-world projects requires resolving cross-module dependencies and global contexts, which are challenges overlooked by existing function-centric methods. We introduce RagVerus, a framework that synergizes retrieval-augmented generation with context-aware prompting to automate proof synthesis for multi-module repositories, achieving a 27% relative improvement on our novel RepoVBench benchmark -- the first repository-level dataset for Verus with 383 proof completion tasks. RagVerus triples proof pass rates on existing benchmarks under constrained language model budgets, demonstrating a scalable and sample-efficient verification.

cs.SE↗

arXiv2024

Decoupling Training-Free Guided Diffusion by ADMM

In this paper, we consider the conditional generation problem by guiding off-the-shelf unconditional diffusion models with differentiable loss functions in a plug-and-play fashion. While previous research has primarily focused on balancing the unconditional diffusion model and the guided loss through a tuned weight hyperparameter, we propose a novel framework that distinctly decouples these two components. Specifically, we introduce two variables ${x}$ and ${z}$, to represent the generated samples governed by the unconditional generation model and the guidance function, respectively. This decoupling reformulates conditional generation into two manageable subproblems, unified by the constraint ${x} = {z}$. Leveraging this setup, we develop a new algorithm based on the Alternating Direction Method of Multipliers (ADMM) to adaptively balance these components. Additionally, we establish the equivalence between the diffusion reverse step and the proximal operator of ADMM and provide a detailed convergence analysis of our algorithm under certain mild assumptions. Our experiments demonstrate that our proposed method ADMMDiff consistently generates high-quality samples while ensuring strong adherence to the conditioning criteria. It outperforms existing methods across a range of conditional generation tasks, including image generation with various guidance and controllable motion synthesis.

cs.CV↗

arXiv2024

Code Repair with LLMs gives an Exploration-Exploitation Tradeoff

Iteratively improving and repairing source code with large language models (LLMs), known as refinement, has emerged as a popular way of generating programs that would be too complex to construct in one shot. Given a bank of test cases, together with a candidate program, an LLM can improve that program by being prompted with failed test cases. But it remains an open question how to best iteratively refine code, with prior work employing simple greedy or breadth-first strategies. We show here that refinement exposes an explore-exploit tradeoff: exploit by refining the program that passes the most test cases, or explore by refining a lesser considered program. We frame this as an arm-acquiring bandit problem, which we solve with Thompson Sampling. The resulting LLM-based program synthesis algorithm is broadly applicable: Across loop invariant synthesis, visual reasoning puzzles, and competition programming problems, we find that our new method can solve more problems using fewer language model calls.

cs.SE↗

arXiv2024

Library Learning Doesn't: The Curious Case of the Single-Use "Library"

Advances in Large Language Models (LLMs) have spurred a wave of LLM library learning systems for mathematical reasoning. These systems aim to learn a reusable library of tools, such as formal Isabelle lemmas or Python programs that are tailored to a family of tasks. Many of these systems are inspired by the human structuring of knowledge into reusable and extendable concepts, but do current methods actually learn reusable libraries of tools? We study two library learning systems for mathematics which both reported increased accuracy: LEGO-Prover and TroVE. We find that function reuse is extremely infrequent on miniF2F and MATH. Our followup ablation experiments suggest that, rather than reuse, self-correction and self-consistency are the primary drivers of the observed performance gains. Our code and data are available at https://github.com/ikb-a/curious-case

cs.LG↗

arXiv2024

A Survey on Deep Learning for Theorem Proving

Theorem proving is a fundamental aspect of mathematics, spanning from informal reasoning in natural language to rigorous derivations in formal systems. In recent years, the advancement of deep learning, especially the emergence of large language models, has sparked a notable surge of research exploring these techniques to enhance the process of theorem proving. This paper presents a comprehensive survey of deep learning for theorem proving by offering (i) a thorough review of existing approaches across various tasks such as autoformalization, premise selection, proofstep generation, and proof search; (ii) an extensive summary of curated datasets and strategies for synthetic data generation; (iii) a detailed analysis of evaluation metrics and the performance of state-of-the-art methods; and (iv) a critical discussion on the persistent challenges and the promising avenues for future exploration. Our survey aims to serve as a foundational reference for deep learning approaches in theorem proving, inspiring and catalyzing further research endeavors in this rapidly growing field. A curated list of papers is available at https://github.com/zhaoyu-li/DL4TP.

cs.AI↗

arXiv2024

Modernizing SMT-Based Type Error Localization

Traditional implementations of strongly-typed functional programming languages often miss the root cause of type errors. As a consequence, type error messages are often misleading and confusing - particularly for students learning such a language. We describe Tyro, a type error localization tool which determines the optimal source of an error for ill-typed programs following fundamental ideas by Pavlinovic et al. : we first translate typing constraints into SMT (Satisfiability Modulo Theories) using an intermediate representation which is more readable than the actual SMT encoding; during this phase we apply a new encoding for polymorphic types. Second, we translate our intermediate representation into an actual SMT encoding and take advantage of recent advancements in off-the-shelf SMT solvers to effectively find optimal error sources for ill-typed programs. Our design maintains the separation of heuristic and search also present in prior and similar work. In addition, our architecture design increases modularity, re-usability, and trust in the overall architecture using an intermediate representation to facilitate the safe generation of the SMT encoding. We believe this design principle will apply to many other tools that leverage SMT solvers. Our experimental evaluation reinforces that the SMT approach finds accurate error sources using both expert-labeled programs and an automated method for larger-scale analysis. Compared to prior work, Tyro lays the basis for large-scale evaluation of error localization techniques, which can be integrated into programming environments and enable us to understand the impact of precise error messages for students in practice.

cs.PL↗

arXiv2024

Assessing Code Generation with Intermediate Languages

Intermediate step methodologies like chain of thoughts (COT) have demonstrated effectiveness in enhancing the performance of Large Language Models (LLMs) on code generation. This study explores the utilization of intermediate languages, including various programming languages, natural language solutions, and pseudo-code, and systematically evaluates their impact on the performance of LLMs in code generation tasks. Our experiments encompass eleven models across the CodeLlama, GPT, and Mistral families, as well as newly released smaller models. Our findings reveal that intermediate languages generally exhibit greater efficacy in larger models that have not yet achieved state-of-the-art performance. Natural language consistently emerges as the most effective intermediate representation across all target languages. However, we observe no universally effective intermediate formal language across different models and target languages. Furthermore, we uncover a weak correlation between the correctness of intermediate solutions and final generation, suggesting that improvements may stem from the chain-of-thought effect rather than language-specific transfer. Interestingly, we discover that for GPT family models, prompting multiple times without explicit self-correction instructions yields performance gains across the studied languages.

cs.SE↗

arXiv2024

APPL: A Prompt Programming Language for Harmonious Integration of Programs and Large Language Model Prompts

Large Language Models (LLMs) have become increasingly capable of handling diverse tasks with the aid of well-crafted prompts and integration of external tools, but as task complexity rises, the workflow involving LLMs can be complicated and thus challenging to implement and maintain. To address this challenge, we propose APPL, A Prompt Programming Language that acts as a bridge between computer programs and LLMs, allowing seamless embedding of prompts into Python functions, and vice versa. APPL provides an intuitive and Python-native syntax, an efficient parallelized runtime with asynchronous semantics, and a tracing module supporting effective failure diagnosis and replaying without extra costs. We demonstrate that APPL programs are intuitive, concise, and efficient through three representative scenarios: Chain-of-Thought with self-consistency (CoT-SC), ReAct tool use agent, and multi-agent chat. Experiments on three parallelizable workflows further show that APPL can effectively parallelize independent LLM calls, with a significant speedup ratio that almost matches the estimation.

cs.AI↗

arXiv2024

Chronosymbolic Learning: Efficient CHC Solving with Symbolic Reasoning and Inductive Learning

Solving Constrained Horn Clauses (CHCs) is a fundamental challenge behind a wide range of verification and analysis tasks. Data-driven approaches show great promise in improving CHC solving without the painstaking manual effort of creating and tuning various heuristics. However, a large performance gap exists between data-driven CHC solvers and symbolic reasoning-based solvers. In this work, we develop a simple but effective framework, "Chronosymbolic Learning", which unifies symbolic information and numerical data points to solve a CHC system efficiently. We also present a simple instance of Chronosymbolic Learning with a data-driven learner and a BMC-styled reasoner. Despite its relative simplicity, experimental results show the efficacy and robustness of our tool. It outperforms state-of-the-art CHC solvers on a dataset consisting of 288 benchmarks, including many instances with non-linear integer arithmetics.

cs.LO↗