arXiv subjects

Sanjit A. Seshia

Publications and source records attributed to Sanjit A. Seshia.

At least 19 recordsLinked to original sources

arXiv2026

Scenario-Based Compositional Statistical Model Checking for Safety Specifications

In safety-critical domains such as autonomous driving, systems must be evaluated across a large number of environment conditions, often represented as composite scenarios built from primitive scenarios. Existing statistical model checking (SMC) approaches analyze each composite scenario independently, requiring many expensive simulations and resulting in substantial redundant computation when scenarios share common structure. This work introduces a scenario-based compositional SMC framework for safety and co-safety specifications, enabling efficient analysis of composite scenarios. Our approach decomposes scenarios into primitives and specifications into sub-specifications, verifies each primitive independently, and composes the resulting statistical estimates using importance sampling and kernel density estimation. Our empirical evaluation shows that the proposed framework can accurately answer verification queries for previously unseen composite scenarios while reducing simulation cost through parallelization and trace reuse.

cs.LO↗

arXiv2026

Any-Body Guard: Universal Safeguarding for Manipulation Policies via Action Masking

Ensuring safety of learning-enabled robotic manipulation across diverse embodiments and tasks still requires significant manual engineering. Existing approaches typically rely on heuristically designed fallback controllers or complex forward invariance assessments. These methods are often too conservative for task success, too computationally expensive for real-time execution, too heuristic to provide useful safety guarantees, or too engineering-heavy to transfer between setups. In this paper, we propose a universal safeguarding approach, X-Safe, which reasons directly in the robot's configuration space to provide formal probabilistic guarantees for collision avoidance. By operating in the configuration space, our method transfers across embodiments while relying solely on an object-based, quasi-static scene representation and a forward kinematics model of the robotic manipulator. Thus, X-Safe provides useful formal safety guarantees without requiring additional data, or engineering effort for different embodiments or scenes. We demonstrate X-Safe for diverse embodiments and policies, both in simulation and on hardware. We observe less degradation in task performance compared to state-of-the-art safeguarding, no collisions on hardware experiments, and empirically corroborate our formal guarantees.

cs.RO↗

arXiv2026

RHO: Your Coding Agent is Secretly a Roboticist

Code-as-Policies (CaP) has shown that large language models (LLMs) can write code to solve robotics tasks by composing perception, planning, and control primitives. Recent CaP systems, however, rely on multi-turn code-generation loops at test time, which is often infeasible for real-time robot control. We introduce Robotics Harness Optimization (RHO), a novel paradigm in which tool-enabled coding agents, at training time, propose and search for interpretable, neurosymbolic multi-file policy repositories (Repositories-as-Policies) that compose these primitives rather than a single prompt, function, or file. RHO searches with reflective feedback from environment reward and execution rather than teleoperation demonstrations. It generalizes to perturbed pick-and-place settings like LIBERO-PRO, where OpenVLA scores 0.0% and $π_{0.5}$ averages 12.83%. Using the same low-level primitives, RHO reaches a 45.0% success rate, 2.5x higher than the strongest multi-turn agentic system, and 3.5x higher than $π_{0.5}$. On Robosuite, RHO sets a new state-of-the-art of 70.0%, exceeding the prior multi-turn record of 68.29% using single-turn execution with no corrective LLM code edits at deployment. When an LLM is used in the control loop, as on RAI's O3DE benchmark, RHO optimizes the deployed agent's multi-file harness of prompts, tools, and control code, improving held-out success from 23.5% to 44.3% with 20% less wall-clock time and 27% fewer tool calls.

cs.RO↗

arXiv2026

ScenicRules: An Autonomous Driving Benchmark with Multi-Objective Specifications and Abstract Scenarios

Developing autonomous driving systems for complex traffic environments requires balancing multiple objectives, such as avoiding collisions, obeying traffic rules, and making efficient progress. In many situations, these objectives cannot be satisfied simultaneously, and explicit priority relations naturally arise. Also, driving rules require context, so it is important to formally model the environment scenarios within which such rules apply. Existing benchmarks for evaluating autonomous vehicles lack such combinations of multi-objective prioritized rules and formal environment models. In this work, we introduce ScenicRules, a benchmark for evaluating autonomous driving systems in stochastic environments under prioritized multi-objective specifications. We first formalize a diverse set of objectives to serve as quantitative evaluation metrics. Next, we design a Hierarchical Rulebook framework that encodes multiple objectives and their priority relations in an interpretable and adaptable manner. We then construct a compact yet representative collection of scenarios spanning diverse driving contexts and near-accident situations, formally modeled in the Scenic language. Experimental results show that our formalized objectives and Hierarchical Rulebooks align well with human driving judgments and that our benchmark effectively exposes agent failures with respect to the prioritized objectives. Our benchmark can be accessed at https://github.com/BerkeleyLearnVerify/ScenicRules/.

cs.RO↗

arXiv2026

Automata-Conditioned Cooperative Multi-Agent Reinforcement Learning

We study learning multi-task, multi-agent policies for cooperative, temporal objectives, under centralized training, decentralized execution. In this setting, using automata to represent tasks assigned to agents enables breaking down a team-level objective into simpler, smaller sub-tasks. However, existing approaches remain sample-inefficient and are limited to the single-task case, requiring retraining policies for each new task. In this work, we present Automata-Conditioned Cooperative Multi-Agent Reinforcement Learning (ACC-MARL), a framework for learning task-conditioned, decentralized team policies. We identify challenges to the feasibility of ACC-MARL, propose solutions, and prove that our approach is optimal. We further show that learned value functions can be used to assign tasks optimally at test time. Experiments demonstrate emergent task-aware, multi-step coordination among agents, such as pressing a button to unlock a door, holding the door, and short-circuiting tasks.

cs.MA↗

arXiv2026

optimize_anything: A Universal API for Optimizing any Text Parameter

Can a single LLM-based optimization system match specialized tools across fundamentally different domains? We show that when optimization problems are formulated as improving a text artifact evaluated by a scoring function, a single AI-based optimization system-supporting single-task search, multi-task search with cross-problem transfer, and generalization to unseen inputs-achieves state-of-the-art results across six diverse tasks. Our system discovers agent architectures that nearly triple Gemini Flash's ARC-AGI accuracy (32.5% to 89.5%), finds scheduling algorithms that cut cloud costs by 40%, generates CUDA kernels where 87% match or beat PyTorch, and outperforms AlphaEvolve's reported circle packing solution (n=26). Ablations across three domains reveal that actionable side information yields faster convergence and substantially higher final scores than score-only feedback, and that multi-task search outperforms independent optimization given equivalent per-problem budget through cross-task transfer, with benefits scaling with the number of related tasks. Together, we show for the first time that text optimization with LLM-based search is a general-purpose problem-solving paradigm, unifying tasks traditionally requiring domain-specific algorithms under a single framework. We open-source optimize\_anything with support for multiple backends as part of the GEPA project at https://github.com/gepa-ai/gepa .

cs.CL↗

arXiv2026

Learning Contextual Runtime Monitors for Safe AI-Based Autonomy

We introduce a novel framework for learning context-aware runtime monitors for AI-based control ensembles. Machine-learning (ML) controllers are increasingly deployed in (autonomous) cyber-physical systems because of their ability to solve complex decision-making tasks. However, their accuracy can degrade sharply in unfamiliar environments, creating significant safety concerns. Traditional ensemble methods aim to improve robustness by averaging or voting across multiple controllers, yet this often dilutes the specialized strengths that individual controllers exhibit in different operating contexts. We argue that, rather than blending controller outputs, a monitoring framework should identify and exploit these contextual strengths. In this paper, we reformulate the design of safe AI-based control ensembles as a contextual monitoring problem. A monitor continuously observes the system's context and selects the controller best suited to the current conditions. To achieve this, we cast monitor learning as a contextual learning task and draw on techniques from contextual multi-armed bandits. Our approach comes with two key benefits: (1) theoretical safety guarantees during controller selection, and (2) improved utilization of controller diversity. We validate our framework in two simulated autonomous driving scenarios, demonstrating significant improvements in both safety and performance compared to non-contextual baselines.

cs.LG↗

arXiv2025

Clinician-Directed Large Language Model Software Generation for Therapeutic Interventions in Physical Rehabilitation

Digital health interventions increasingly deliver home exercise programs via sensor-equipped devices such as smartphones, enabling remote monitoring of adherence and performance. However, current software is usually authored before clinical encounters as libraries of modules for broad impairment categories. At the point of care, clinicians can only choose from these modules and adjust a few parameters (for example, duration or repetitions). As a result, individual limitations, goals, and environmental constraints are often not reflected, limiting personalization and benefit. We propose a paradigm in which large language models (LLMs) act as constrained translators that convert clinicians' exercise prescriptions into intervention software. Clinicians remain the decision makers: they design exercises during the encounter, tailored to each patient's impairments, goals, and environment, and the LLM generates matching software. We conducted a prospective single-arm feasibility study with 20 licensed physical and occupational therapists who created 40 individualized upper extremity programs for a standardized patient; 100% of prescriptions were translated into executable software, compared with 55% under a representative template-based digital health intervention (p < 0.01). LLM-generated software correctly delivered 99.7% of instructions and monitored performance with 88.4% accuracy (95% confidence interval, 0.843-0.915). Overall, 90% of therapists judged the system safe for patient interaction and 75% expressed willingness to adopt it in practice. To our knowledge, this is the first prospective evaluation of clinician-directed intervention software generation with an LLM in health care, demonstrating feasibility and motivating larger trials in real patient populations.

cs.HC↗

arXiv2025

Robust and Diverse Multi-Agent Learning via Rational Policy Gradient

Adversarial optimization algorithms that explicitly search for flaws in agents' policies have been successfully applied to finding robust and diverse policies in multi-agent settings. However, the success of adversarial optimization has been largely limited to zero-sum settings because its naive application in cooperative settings leads to a critical failure mode: agents are irrationally incentivized to self-sabotage, blocking the completion of tasks and halting further learning. To address this, we introduce Rationality-preserving Policy Optimization (RPO), a formalism for adversarial optimization that avoids self-sabotage by ensuring agents remain rational--that is, their policies are optimal with respect to some possible partner policy. To solve RPO, we develop Rational Policy Gradient (RPG), which trains agents to maximize their own reward in a modified version of the original game in which we use opponent shaping techniques to optimize the adversarial objective. RPG enables us to extend a variety of existing adversarial optimization algorithms that, no longer subject to the limitations of self-sabotage, can find adversarial examples, improve robustness and adaptability, and learn diverse policies. We empirically validate that our approach achieves strong performance in several popular cooperative and general-sum environments. Our project page can be found at https://rational-policy-gradient.github.io.

cs.AI↗

arXiv2025

ScenicProver: A Framework for Compositional Probabilistic Verification of Learning-Enabled Systems

Full verification of learning-enabled cyber-physical systems (CPS) has long been intractable due to challenges including black-box components and complex real-world environments. Existing tools either provide formal guarantees for limited types of systems or test the system as a monolith, but no general framework exists for compositional analysis of learning-enabled CPS using varied verification techniques over complex real-world environments. This paper introduces ScenicProver, a verification framework that aims to fill this gap. Built upon the Scenic probabilistic programming language, the framework supports: (1) compositional system description with clear component interfaces, ranging from interpretable code to black boxes; (2) assume-guarantee contracts over those components using an extension of Linear Temporal Logic containing arbitrary Scenic expressions; (3) evidence generation through testing, formal proofs via Lean 4 integration, and importing external assumptions; (4) systematic combination of generated evidence using contract operators; and (5) automatic generation of assurance cases tracking the provenance of system-level guarantees. We demonstrate the framework's effectiveness through a case study on an autonomous vehicle's automatic emergency braking system with sensor fusion. By leveraging manufacturer guarantees for radar and laser sensors and focusing testing efforts on uncertain conditions, our approach enables stronger probabilistic guarantees than monolithic testing with the same computational budget.

cs.LO↗

arXiv2025

GRAID: Enhancing Spatial Reasoning of VLMs Through High-Fidelity Data Generation

Vision Language Models (VLMs) achieve strong performance on many vision-language tasks but often struggle with spatial reasoning$\unicode{x2014}$a prerequisite for many applications. Empirically, we find that a dataset produced by a current training data generation pipeline has a 57.6% human validation rate. These rates stem from current limitations: single-image 3D reconstruction introduces cascading modeling errors and requires wide answer tolerances, while caption-based methods require hyper-detailed annotations and suffer from generative hallucinations. We present GRAID, built on the key insight that qualitative spatial relationships can be reliably determined from 2D geometric primitives alone. By operating exclusively on 2D bounding boxes from standard object detectors, GRAID avoids both 3D reconstruction errors and generative hallucinations, resulting in datasets that are of higher quality than existing tools that produce similar datasets as validated by human evaluations. We apply our framework to the BDD100k, NuImages, and Waymo datasets, generating over 8.5 million high-quality VQA pairs creating questions spanning spatial relations, counting, ranking, and size comparisons. We evaluate one of the datasets and find it achieves 91.16% human-validated accuracy$\unicode{x2014}$compared to 57.6% on a dataset generated by recent work. Critically, we demonstrate that when trained on GRAID data, models learn spatial reasoning concepts that generalize: models fine-tuned on 6 question types improve on over 10 held-out types, with accuracy gains of 47.5% on BDD and 37.9% on NuImages for Llama 3.2B 11B, and when trained on all questions types, achieve improvements on several existing benchmarks such as BLINK. The GRAID framework, datasets, and additional information can be found $\href{this https URL}{here}$.

cs.CV↗

arXiv2025

Learning Affordances at Inference-Time for Vision-Language-Action Models

Solving complex real-world control tasks often takes multiple tries: if we fail at first, we reflect on what went wrong, and change our strategy accordingly to avoid making the same mistake. In robotics, Vision-Language-Action models (VLAs) offer a promising path towards solving complex control tasks, but lack the ability to contextually and dynamically readjust behavior when they fail to accomplish a task. In this work, we introduce Learning from Inference-Time Execution (LITEN), which connects a VLA low-level policy to a high-level VLM that conditions on past experiences by including them in-context, allowing it to learn the affordances and capabilities of the low-level VLA. Our approach iterates between a reasoning phase that generates and executes plans for the low-level VLA, and an assessment phase that reflects on the resulting execution and draws useful conclusions to be included in future reasoning contexts. Unlike similar approaches to self-refinement in non-robotics domains, LITEN must reflect on unstructured real-world robot trajectories (e.g., raw videos), which requires structured guiderails during assessment. Our experimental results demonstrate LITEN is able to effectively learn from past experience to generate plans that use high-affordance instructions to accomplish long-horizon tasks.

cs.RO↗

arXiv2025

PolyVer: A Compositional Approach for Polyglot System Modeling and Verification

Several software systems are polyglot; that is, they comprise programs implemented in a combination of programming languages. Verifiers that directly run on mainstream programming languages are currently customized for single languages. Thus, to verify polyglot systems, one usually translates them into a common verification language or formalism on which the verifier runs. In this paper, we present an alternative approach, PolyVer, which employs abstraction, compositional reasoning, and synthesis to directly perform polyglot verification. PolyVer constructs a formal model of the original polyglot system as a transition system where the update functions associated with transitions are implemented in target languages such as C or Rust. To perform verification, PolyVer then connects a model checker for transition systems with language-specific verifiers (e.g., for C or Rust) using pre/post-condition contracts for the update functions. These contracts are automatically generated by synthesis oracles based on syntax-guided synthesis or large language models (LLMs), and checked by the language-specific verifiers. The contracts form abstractions of the update functions using which the model checker verifies the overall system-level property on the polyglot system model. PolyVer iterates between counterexample-guided abstraction-refinement (CEGAR) and counterexample-guided inductive synthesis (CEGIS) until the property is verified or a true system-level counterexample is found. We demonstrate the utility of PolyVer for verifying programs in the Lingua Franca polyglot language using the UCLID5 model checker connected with the CBMC and Kani verifiers for C and Rust respectively.

cs.PL↗

arXiv2025

$L^*LM$: Learning Automata from Examples using Natural Language Oracles

Expert demonstrations have proven an easy way to indirectly specify complex tasks. Recent algorithms even support extracting unambiguous formal specifications, e.g. deterministic finite automata (DFA), from demonstrations. Unfortunately, these techniques are generally not sample efficient. In this work, we introduce $L^*LM$, an algorithm for learning DFAs from both demonstrations and natural language. Due to the expressivity of natural language, we observe a significant improvement in the data efficiency of learning DFAs from expert demonstrations. Technically, $L^*LM$ leverages large language models to answer membership queries about the underlying task. This is then combined with recent techniques for transforming learning from demonstrations into a sequence of labeled example learning problems. In our experiments, we observe the two modalities complement each other, yielding a powerful few-shot learner.

cs.LG↗

arXiv2025

Learning Formal Specifications from Membership and Preference Queries

Active learning is a well-studied approach to learning formal specifications, such as automata. In this work, we extend active specification learning by proposing a novel framework that strategically requests a combination of membership labels and pair-wise preferences, a popular alternative to membership labels. The combination of pair-wise preferences and membership labels allows for a more flexible approach to active specification learning, which previously relied on membership labels only. We instantiate our framework in two different domains, demonstrating the generality of our approach. Our results suggest that learning from both modalities allows us to robustly and conveniently identify specifications via membership and preferences.

cs.FL↗

arXiv2025

Provably Correct Automata Embeddings for Optimal Automata-Conditioned Reinforcement Learning

Automata-conditioned reinforcement learning (RL) has given promising results for learning multi-task policies capable of performing temporally extended objectives given at runtime, done by pretraining and freezing automata embeddings prior to training the downstream policy. However, no theoretical guarantees were given. This work provides a theoretical framework for the automata-conditioned RL problem and shows that it is probably approximately correct learnable. We then present a technique for learning provably correct automata embeddings, guaranteeing optimal multi-task policy learning. Our experimental evaluation confirms these theoretical results.

cs.LG↗

arXiv2025

MRTA-Sim: A Modular Simulator for Multi-Robot Allocation, Planning, and Control in Open-World Environments

This paper introduces MRTA-Sim, a Python/ROS2/Gazebo simulator for testing approaches to Multi-Robot Task Allocation (MRTA) problems on simulated robots in complex, indoor environments. Grid-based approaches to MRTA problems can be too restrictive for use in complex, dynamic environments such in warehouses, department stores, hospitals, etc. However, approaches that operate in free-space often operate at a layer of abstraction above the control and planning layers of a robot and make an assumption on approximate travel time between points of interest in the system. These abstractions can neglect the impact of the tight space and multi-agent interactions on the quality of the solution. Therefore, MRTA solutions should be tested with the navigation stacks of the robots in mind, taking into account robot planning, conflict avoidance between robots, and human interaction and avoidance. This tool connects the allocation output of MRTA solvers to individual robot planning using the NAV2 stack and local, centralized multi-robot deconfliction using Control Barrier Function-Quadrtic Programs (CBF-QPs), creating a platform closer to real-world operation for more comprehensive testing of these approaches. The simulation architecture is modular so that users can swap out methods at different levels of the stack. We show the use of our system with a Satisfiability Modulo Theories (SMT)-based approach to dynamic MRTA on a fleet of indoor delivery robots.

cs.RO↗

arXiv2025

LTL-Constrained Policy Optimization with Cycle Experience Replay

Linear Temporal Logic (LTL) offers a precise means for constraining the behavior of reinforcement learning agents. However, in many settings where both satisfaction and optimality conditions are present, LTL is insufficient to capture both. Instead, LTL-constrained policy optimization, where the goal is to optimize a scalar reward under LTL constraints, is needed. This constrained optimization problem proves difficult in deep Reinforcement Learning (DRL) settings, where learned policies often ignore the LTL constraint due to the sparse nature of LTL satisfaction. To alleviate the sparsity issue, we introduce Cycle Experience Replay (CyclER), a novel reward shaping technique that exploits the underlying structure of the LTL constraint to guide a policy towards satisfaction by encouraging partial behaviors compliant with the constraint. We provide a theoretical guarantee that optimizing CyclER will achieve policies that satisfy the LTL constraint with near-optimal probability. We evaluate CyclER in three continuous control domains. Our experimental results show that optimizing CyclER in tandem with the existing scalar reward outperforms existing reward-shaping methods at finding performant LTL-satisfying policies.

cs.LG↗