Multi-Level Intermediate Representation (MLIR) underlies modern ML compiler infrastructure (TensorFlow, JAX/StableHLO, PyTorch Inductor, IREE), yet appears only in trace amounts in code-LM pretraining corpora. MLIR is also extensible by design: new dialects ship per application domain, so a fine-tuned model per dialect...
A benign image paired with a benign caption can become harmful only in combination, making detection fundamentally a cross-modal reasoning problem. Existing multimodal harm benchmarks evaluate only classification accuracy, leaving unclear whether models identify the cross-modal mechanism that produces harm. We introduc...
Anisha Saha, Varsha Suresh, Teodora Kamova et al.· 0 citations
Ranking systems used for decision support should expose evidence that can be audited independently of the model's ranking scores. We study a closed-roster setting in which a policy jointly emits a Top-$K$ ranking and candidate-linked, span-grounded evidence bundles. Evidence-Coupled Policy Optimization (ECPO) combines...
Miaobo Hu, Shuhao Hu, BoKun Wang et al.· 0 citations
Randomized experiments are the standard for measuring causal effects that guide decisions, but are costly. Simulators calibrated to historical data offer a cheaper alternative, but their policies can become suboptimal under drift, a change in dynamics after calibration. We study if, when, and how to collect data for re...
Harsh Parikh, Gabriel Levin-Konigsberg, Sohrab Andaz et al.· 0 citations
Reach audiences
Advertise in front of researchers, engineers, and readers.
Cooperative multi-agent reinforcement learning (MARL) benchmarks commonly emphasize aggregate outcomes such as return, success rate, or completion time. While essential, these metrics can obscure how agents coordinate and fail to distinguish policies that achieve similar outcomes through different interaction processes...
Maria Ana Cardei, Matthew Landers, Afsaneh Doryab· 0 citations
Code assistants are increasingly utilized in test-driven software development, yet the theoretical mechanisms behind their environment-interaction strategies remain underexplored. We provide a probabilistic framework for two dominant paradigms: code selection after generation using the execution environment, and code g...
Nicolas Menet, Michael Hersche, Andreas Krause et al.· 0 citations
Large language models (LLMs) are increasingly employed for decision-support across multiple domains. We ask whether these models favor artificial intelligence (AI) itself. We examine this question through three experiments. First, in open-ended advice across investment, study, career, and startup domains, AI is promine...
Benaya Trabelsi, Jonathan Shaki, Sarit Kraus· 0 citations
Modern data processing workflows frequently encounter ragged data: collections with variable-length elements that arise naturally in domains like natural language processing, scientific measurements, and autonomous AI agents. Existing workflow engines lack native support for tracking the shapes and dependencies inheren...
Sungbin Moon, Jiho Park, Suyoung Hwang et al.· 0 citations
Large language models represent text as hidden numerical patterns, but what information these representations actually contain remains a mystery. Can we recover the original text from just one of them? In this paper, we introduce Rep2Text, a method designed to answer this question by reconstructing text from a single h...
Haiyan Zhao, Zirui He, Yiming Tang et al.· 0 citations
Handling missing values is a common challenge in biostatistical analyses, typically addressed by imputation methods. We propose a novel, fast, and easy-to-use imputation method called missing value imputation with adversarial random forests (MissARF), based on generative machine learning, that provides both single and...
Pegah Golchian, Jan Kapar, David S. Watson et al.· 0 citations
This paper focuses on a sub-domain of XAI, called eXplainable Reinforcement Learning (XRL), which aims to explain the actions of an agent that has learned by reinforcement learning, and proposes an intuitive taxonomy based on two questions, based on the target that the method explains.
Post-training plays a crucial role in refining and aligning large language models to meet specific tasks and human preferences. While recent advancements in post-training techniques, such as Group Relative Policy Optimization (GRPO), leverage increased sampling with relative reward scoring to achieve superior performan...
Kaichen Zhang, Yuzhong Hong, Junwei Bao et al.· 0 citations
