Generating diverse responses from large language models (LLMs) is crucial for applications such as planning/search and synthetic data generation, where diversity provides distinct answers across generations. Prior approaches rely on increasing temperature to increase diversity. However, contrary to popular belief, we show not only does this approach produce lower quality individual generations as temperature increases, but it depends on model's next-token probabilities being similar to the true distribution of answers. We propose SimpleStrat, an alternative approach that uses the language model itself to partition the space into strata. At inference, a random stratum is selected and a sample drawn from within the strata. To measure diversity, we introduce CoverageQA, a dataset of underspecified questions with multiple equally plausible answers, and assess diversity by measuring KL Divergence between the output distribution and uniform distribution over valid ground truth answers. As computing probability per response/solution for proprietary models is infeasible, we measure recall on ground truth solutions. Our evaluation show using SimpleStrat achieves higher recall by 0.05 compared to GPT-4o and 0.36 average reduction in KL Divergence compared to Llama 3.
Reliable estimation of treatment effects from observational data is important in many disciplines such as medicine. However, estimation is challenging when unconfoundedness as a standard assumption in the causal inference literature is violated. In this work, we leverage arbitrary (potentially high-dimensional) instruments to estimate bounds on the conditional average treatment effect (CATE). Our contributions are three-fold: (1) We propose a novel approach for partial identification through a mapping of instruments to a discrete representation space so that we yield valid bounds on the CATE. This is crucial for reliable decision-making in real-world applications. (2) We derive a two-step procedure that learns tight bounds using a tailored neural partitioning of the latent instrument space. As a result, we avoid instability issues due to numerical approximations or adversarial training. Furthermore, our procedure aims to reduce the estimation variance in finite-sample settings to yield more reliable estimates. (3) We show theoretically that our procedure obtains valid bounds while reducing estimation variance. We further perform extensive experiments to demonstrate the effectiveness across various settings. Overall, our procedure offers a novel path for practitioners to make use of potentially high-dimensional instruments (e.g., as in Mendelian randomization).
Algorithmic Recourse (AR) aims to provide users with actionable steps to overturn unfavourable decisions made by machine learning predictors. However, these actions often take time to implement (e.g., getting a degree can take years), and their effects may vary as the world evolves. Thus, it is natural to ask for recourse that remains valid in a dynamic environment. In this paper, we study the robustness of algorithmic recourse over time by casting the problem through the lens of causality. We demonstrate theoretically and empirically that (even robust) causal AR methods can fail over time, except in the -- unlikely -- case that the world is stationary. Even more critically, unless the world is fully deterministic, counterfactual AR cannot be solved optimally. To account for this, we propose a simple yet effective algorithm for temporal AR that explicitly accounts for time under the assumption of having access to an estimator approximating the stochastic process. Our simulations on synthetic and realistic datasets show how considering time produces more resilient solutions to potential trends in the data distribution.
Designing ligand-binding proteins, such as enzymes and biosensors, is essential in bioengineering and protein biology. One critical step in this process involves designing protein pockets, the protein interface binding with the ligand. Current approaches to pocket generation often suffer from time-intensive physical computations or template-based methods, as well as compromised generation quality due to the overlooking of domain knowledge. To tackle these challenges, we propose PocketFlow, a generative model that incorporates protein-ligand interaction priors based on flow matching. During training, PocketFlow learns to model key types of protein-ligand interactions, such as hydrogen bonds. In the sampling, PocketFlow leverages multi-granularity guidance (overall binding affinity and interaction geometry constraints) to facilitate generating high-affinity and valid pockets. Extensive experiments show that PocketFlow outperforms baselines on multiple benchmarks, e.g., achieving an average improvement of 1.29 in Vina Score and 0.05 in scRMSD. Moreover, modeling interactions make PocketFlow a generalized generative model across multiple ligand modalities, including small molecules, peptides, and RNA.
Economists are often interested in functions of multiple causal effects, a leading example of which is evaluating a policy's cost-effectiveness. The benefits and costs might be captured by multiple causal effects and aggregated into a scalar measure of cost-effectiveness. Oftentimes, the microdata underlying these estimates is inaccessible; only published estimates and their corresponding standard errors are available. We provide a method to conduct inference on non-linear functions of causal effects when the only information available is the point estimates and their standard errors. We apply our method to inference for the Marginal Value of Public Funds (MVPF) of government policies.
Abstraction-based control design is a promising approach for ensuring safety-critical control of complex cyber-physical systems. A key aspect of this methodology is the relation between the original and abstract systems, which ensures that the abstract controller can be transformed into a valid controller for the original system through a concretization procedure. In this paper, we provide a comprehensive and systematic framework that characterizes various simulation relations, through their associated concretization procedures. We introduce the concept of augmented system, which universally enables a feedback refinement relation with the abstract system. This augmented system encapsulates the specific characteristics of each simulation relation within an interface, enabling a plug-and-play control architecture. Our results demonstrate that the existence of a particular simulation relation between the concrete and abstract systems is equivalent to the implementability of a specific control architecture, which depends on the considered simulation relation. This allows us to introduce new types of relations, and to establish the advantages and drawbacks of different relations, which we exhibit through detailed examples.
In this paper, we develop a framework for the automatic taxiing of aircraft between hangar and take-off given a graph-based model of an airport. We implement a high-level path-planning algorithm that models taxiway intersections as nodes in an undirected graph, algorithmically constructs a directed graph according to the physical limitations of the aircraft, and finds the shortest valid taxi path through the directed graph using Dijkstra's algorithm. We then use this shortest path to construct a reference trajectory for the aircraft to follow that considers the turning capabilities of a given aircraft. Using high-order control barrier functions (HOCBFs), we construct safety conditions for multi-obstacle avoidance and safe reference tracking for simple 2D unicycle dynamics with acceleration control inputs. We then use these safety conditions to design an MPC-CBF framework that tracks the reference trajectory while adhering to the safety constraints. We compare the performance of our MPC-CBF controller with a PID-CBF control method via simulations.
The Gaussian process (GP) is a Bayesian nonparametric paradigm that is widely adopted for uncertainty quantification (UQ) in a number of safety-critical applications, including robotics, healthcare, as well as surveillance. The consistency of the resulting uncertainty values however, hinges on the premise that the learning function conforms to the properties specified by the GP model, such as smoothness, periodicity and more, which may not be satisfied in practice, especially with data arriving on the fly. To combat against such model mis-specification, we propose to wed the GP with the prevailing conformal prediction (CP), a distribution-free post-processing framework that produces it prediction sets with a provably valid coverage under the sole assumption of data exchangeability. However, this assumption is usually violated in the online setting, where a prediction set is sought before revealing the true label. To ensure long-term coverage guarantee, we will adaptively set the key threshold parameter based on the feedback whether the true label falls inside the prediction set. Numerical results demonstrate the merits of the online GP-CP approach relative to existing alternatives in the long-term coverage performance.
While large code language models have made significant strides in AI-assisted coding tasks, there are growing concerns about privacy challenges. The user code is transparent to the cloud LLM service provider, inducing risks of unauthorized training, reading, and execution of the user code. In this paper, we propose CodeCipher, a novel method that perturbs privacy from code while preserving the original response from LLMs. CodeCipher transforms the LLM's embedding matrix so that each row corresponds to a different word in the original matrix, forming a token-to-token confusion mapping for obfuscating source code. The new embedding matrix is optimized by minimizing the task-specific loss function. To tackle the challenge of the discrete and sparse nature of word vector spaces, CodeCipher adopts a discrete optimization strategy that aligns the updated vector to the nearest valid token in the vocabulary before each gradient update. We demonstrate the effectiveness of our approach on three AI-assisted coding tasks including code completion, summarization, and translation. Results show that our model successfully confuses the privacy in source code while preserving the original LLM's performance.
Conformal prediction (CP), a distribution-free uncertainty quantification (UQ) framework, reliably provides valid predictive inference for black-box models. CP constructs prediction sets that contain the true output with a specified probability. However, modern data science diverse modalities, along with increasing data and model complexity, challenge traditional CP methods. These developments have spurred novel approaches to address evolving scenarios. This survey reviews the foundational concepts of CP and recent advancements from a data-centric perspective, including applications to structured, unstructured, and dynamic data. We also discuss the challenges and opportunities CP faces in large-scale data and models.
Scientific discovery plays a pivotal role in advancing human society, and recent progress in large language models (LLMs) suggests their potential to accelerate this process. However, it remains unclear whether LLMs can autonomously generate novel and valid hypotheses in chemistry. In this work, we investigate whether LLMs can discover high-quality chemistry hypotheses given only a research background-comprising a question and/or a survey-without restriction on the domain of the question. We begin with the observation that hypothesis discovery is a seemingly intractable task. To address this, we propose a formal mathematical decomposition grounded in a fundamental assumption: that most chemistry hypotheses can be composed from a research background and a set of inspirations. This decomposition leads to three practical subtasks-retrieving inspirations, composing hypotheses with inspirations, and ranking hypotheses - which together constitute a sufficient set of subtasks for the overall scientific discovery task. We further develop an agentic LLM framework, MOOSE-Chem, that is a direct implementation of this mathematical decomposition. To evaluate this framework, we construct a benchmark of 51 high-impact chemistry papers published and online after January 2024, each manually annotated by PhD chemists with background, inspirations, and hypothesis. The framework is able to rediscover many hypotheses with high similarity to the groundtruth, successfully capturing the core innovations-while ensuring no data contamination since it uses an LLM with knowledge cutoff date prior to 2024. Finally, based on LLM's surprisingly high accuracy on inspiration retrieval, a task with inherently out-of-distribution nature, we propose a bold assumption: that LLMs may already encode latent scientific knowledge associations not yet recognized by humans.
The complexity of the cardiovascular system needs to be accurately reproduced in order to promptly acknowledge health conditions; to this aim, advanced multifidelity and multiphysics numerical models are crucial. On one side, Full Order Models (FOMs) deliver accurate hemodynamic assessments, but their high computational demands hinder their real-time clinical application. In contrast, Reduced Order Models (ROMs) provide more efficient yet accurate solutions, essential for personalized healthcare and timely clinical decision-making. In this work, we explore the application of computational fluid dynamics (CFD) in cardiovascular medicine by integrating FOMs with ROMs for predicting the risk of aortic aneurysm growth and rupture. Wall Shear Stress (WSS) and the Oscillatory Shear Index (OSI), sampled at different growth stages of the thoracic aortic aneurysm, are predicted by means of Graph Neural Networks (GNNs). GNNs exploit the natural graph structure of the mesh obtained by the Finite Volume (FV) discretization, taking into account the spatial local information, regardless of the dimension of the input graph. Our experimental validation framework yields promising results, confirming our method as a valid alternative that overcomes the curse of dimensionality.
Resolving knowledge conflicts is a crucial challenge in Question Answering (QA) tasks, as the internet contains numerous conflicting facts and opinions. While some research has made progress in tackling ambiguous settings where multiple valid answers exist, these approaches often neglect to provide source citations, leaving users to evaluate the factuality of each answer. On the other hand, existing work on citation generation has focused on unambiguous settings with single answers, failing to address the complexity of real-world scenarios. Despite the importance of both aspects, no prior research has combined them, leaving a significant gap in the development of QA systems. In this work, we bridge this gap by proposing the novel task of QA with source citation in ambiguous settings, where multiple valid answers exist. To facilitate research in this area, we create a comprehensive framework consisting of: (1) five novel datasets, obtained by augmenting three existing reading comprehension datasets with citation meta-data across various ambiguous settings, such as distractors and paraphrasing; (2) the first ambiguous multi-hop QA dataset featuring real-world, naturally occurring contexts; (3) two new metrics to evaluate models' performances; and (4) several strong baselines using rule-based, prompting, and finetuning approaches over five large language models. We hope that this new task, datasets, metrics, and baselines will inspire the community to push the boundaries of QA research and develop more trustworthy and interpretable systems.
Answer verification identifies correct solutions among candidates generated by large language models (LLMs). Current approaches typically train verifier models by labeling solutions as correct or incorrect based solely on whether the final answer matches the gold answer. However, this approach neglects any flawed rationale in the solution yielding the correct answer, undermining the verifier's ability to distinguish between sound and flawed rationales. We empirically show that in StrategyQA, only 19% of LLM-generated solutions with correct answers have valid rationales, thus leading to an unreliable verifier. Furthermore, we demonstrate that training a verifier on valid rationales significantly improves its ability to distinguish valid and flawed rationale. To make a better verifier without extra human supervision, we introduce REPS (Rationale Enhancement through Pairwise Selection), a method for selecting valid rationales from candidates by iteratively applying pairwise self-evaluation using the same LLM that generates the solutions. Verifiers trained on solutions selected by REPS outperform those trained using conventional training methods on three reasoning benchmarks (ARC-Challenge, DROP, and StrategyQA). Our results suggest that training reliable verifiers requires ensuring the validity of rationales in addition to the correctness of the final answers, which would be critical for models assisting humans in solving complex reasoning tasks.
We present TANGO, a framework for generating co-speech body-gesture videos. Given a few-minute, single-speaker reference video and target speech audio, TANGO produces high-fidelity videos with synchronized body gestures. TANGO builds on Gesture Video Reenactment (GVR), which splits and retrieves video clips using a directed graph structure - representing video frames as nodes and valid transitions as edges. We address two key limitations of GVR: audio-motion misalignment and visual artifacts in GAN-generated transition frames. In particular, (i) we propose retrieving gestures using latent feature distance to improve cross-modal alignment. To ensure the latent features could effectively model the relationship between speech audio and gesture motion, we implement a hierarchical joint embedding space (AuMoCLIP); (ii) we introduce the diffusion-based model to generate high-quality transition frames. Our diffusion model, Appearance Consistent Interpolation (ACInterp), is built upon AnimateAnyone and includes a reference motion module and homography background flow to preserve appearance consistency between generated and reference videos. By integrating these components into the graph-based retrieval framework, TANGO reliably produces realistic, audio-synchronized videos and outperforms all existing generative and retrieval methods. Our codes and pretrained models are available: \url{https://pantomatrix.github.io/TANGO/}
Automatic program repair at project level may open yet to be seen opportunities in various fields of human activity. Since the SWE-Bench challenge was presented, we have seen numerous of solutions. Patch generation is a part of program repair, and test suite-based conversational patch generation has proven its effectiveness. However, the potential of conversational patch generation has not yet specifically estimated on SWE-Bench. This study reports experimental results aimed at evaluating the individual effectiveness of conversational patch generation on problems from SWE-Bench. The experiments show that a simple conversational pipeline based on LLaMA 3.1 70B can generate valid patches in 47\% of cases, which is comparable to the state-of-the-art in program repair on SWE-Bench.
This paper introduces an innovative approach to analyzing and improving multi-hop reasoning in AI systems by drawing inspiration from Hamiltonian mechanics. We propose a novel framework that maps reasoning chains in embedding spaces to Hamiltonian systems, allowing us to leverage powerful analytical tools from classical physics. Our method defines a Hamiltonian function that balances the progression of reasoning (kinetic energy) against the relevance to the question at hand (potential energy). Using this framework, we analyze a large dataset of reasoning chains from a multi-hop question-answering task, revealing intriguing patterns that distinguish valid from invalid reasoning. We show that valid reasoning chains have lower Hamiltonian energy and move in ways that make the best trade-off between getting more information and answering the right question. Furthermore, we demonstrate the application of this framework to steer the creation of more efficient reasoning algorithms within AI systems. Our results not only provide new insights into the nature of valid reasoning but also open up exciting possibilities for physics-inspired approaches to understanding and improving artificial intelligence.
We introduce G2T-LLM, a novel approach for molecule generation that uses graph-to-tree text encoding to transform graph-based molecular structures into a hierarchical text format optimized for large language models (LLMs). This encoding converts complex molecular graphs into tree-structured formats, such as JSON and XML, which LLMs are particularly adept at processing due to their extensive pre-training on these types of data. By leveraging the flexibility of LLMs, our approach allows for intuitive interaction using natural language prompts, providing a more accessible interface for molecular design. Through supervised fine-tuning, G2T-LLM generates valid and coherent chemical structures, addressing common challenges like invalid outputs seen in traditional graph-based methods. While LLMs are computationally intensive, they offer superior generalization and adaptability, enabling the generation of diverse molecular structures with minimal task-specific customization. The proposed approach achieved comparable performances with state-of-the-art methods on various benchmark molecular generation datasets, demonstrating its potential as a flexible and innovative tool for AI-driven molecular design.
Text-to-SQLs enables non-expert users to effortlessly retrieve desired information from relational databases using natural language queries. While recent advancements, particularly with Large Language Models (LLMs) like GPT and T5, have shown impressive performance on large-scale benchmarks such as BIRD, current state-of-the-art (SOTA) LLM-based Text-to-SQLs models often require significant efforts to develop auxiliary tools like SQL classifiers to achieve high performance. This paper proposed a novel approach that only needs SQL Quality Measurement to enhance LLMs-based Text-to-SQLs performance. It establishes a SQL quality evaluation mechanism to assess the generated SQL queries against predefined criteria and actual database responses. This feedback loop enables continuous learning and refinement of model outputs based on both syntactic correctness and semantic accuracy. The proposed method undergoes comprehensive validation on the BIRD benchmark, assessing Execution Accuracy (EX) and Valid Efficiency Score (VES) across various Text-to-SQLs difficulty levels. Experimental results reveal competitive performance in both EX and VES compared to SOTA models like GPT4 and T5.
We consider the problem of safely exploring a static and unknown environment while learning valid control barrier functions (CBFs) from sensor data. Existing works either assume known environments, target specific dynamics models, or use a-priori valid CBFs, and are thus limited in their safety guarantees for general systems during exploration. We present a method for safely exploring the unknown environment by incrementally composing a global CBF from locally-learned CBFs. The challenge here is that local CBFs may not have well-defined end behavior outside their training domain, i.e. local CBFs may be positive (indicating safety) in regions where no training data is available. We show that well-defined end behavior can be obtained when local CBFs are parameterized by compactly-supported radial basis functions. For learning local CBFs, we collect sensor data, e.g. LiDAR capturing obstacles in the environment, and augment it with simulated data from a safe oracle controller. Our work complements recent efforts to learn CBFs from safe demonstrations -- where learned safe sets are limited to their training domains -- by demonstrating how to grow the safe set over time as more data becomes available. We evaluate our approach on two simulated systems, where our method successfully explores an unknown environment while maintaining safety throughout the entire execution.