Artificial Intelligence · Cross-listed from cs.CV
PACE: Progressive Angular-to-Norm Contrastive Embedding
Yanping Li, Wei Zhou, Yawen Liu et al. (+7 more)
Multimodal embedding models encode heterogeneous inputs into a shared
embedding space, enabling efficient similarity computation across modalities and tasks. Most existing methods optimize cosine-based contrastive objectives, which promote stable training but restrict semantic compatibility to angular geometry, precluding
embedding norms from serving as an additional semantic signal. However, directly optimizing the more expressive dot-product similarity, which leverages both angular and norm information, underperforms cosine-based training and exhibits unstable training dynamics. We attribute this discrepancy to premature optimization-space expansion, manifested as angular–norm
entanglement and directional anisotropy in the representation space and further compounded by full-parameter fine-tuning. In this paper, we propose PACE, a two-stage framework that progressively expands both the representation and trainable parameter spaces. Stage I combines cosine-based objective with low-rank adaptation to establish a reliable angular geometry within constrained optimization spaces. Stage II switches to dot-product similarity and full-parameter fine-tuning, enabling
embedding directions and norms to jointly encode semantic information. We further introduce Focal
Embedding Loss, a confidence-adaptive objective that downweights queries with high positive
retrieval confidence while emphasizing ambiguous queries with competitive negatives. Experiments across multiple backbone scales and diverse
multimodal embedding tasks consistently validate the effectiveness of PACE.
Read the full abstract ▾
All authors Yanping Li, Wei Zhou, Yawen Liu, Yibo Wang, Ke Zhu, Guangda Huzhang, Qing-Guo Chen, Zhao Xu, Jun Zhang, Wei Wei
Artificial Intelligence
HazardAuditor: From Executable Threats to Safer Computer-Use Agents
Yunhao Feng, Ruixiao Lin, Ming Wen et al. (+5 more)
Computer-use
agents increasingly interact with browsers, terminals, file systems, and external services, introducing safety risks that emerge through runtime behavior rather than generated content alone. Existing guard models target static
prompts and responses and are poorly suited to
agent execution; existing executable safety platforms produce evaluation verdicts rather than the normalized supervision a guard model needs to learn across heterogeneous
agent frameworks. We introduce HazardAuditor, an execution-grounded framework that closes both gaps. Its infrastructure runs heterogeneous
agents (Claude Code, Codex, Hermes, and OpenClaw) in controlled environments and normalizes their interactions into a canonical event representation for cross-framework supervision. We further observe that
token-level post-training objectives create a structural mismatch for generative guards, causing longer rationales to dominate gradient updates. Guard Policy Optimization (GuardPO) addresses this by converting deterministic safety outcomes into sequence-level advantages and normalizing rationale and verdict regions, making the safety decision the effective unit of optimization. Across multiple
benchmarks and heterogeneous computer-use systems, HazardAuditor improves accuracy by up to 16.5 percentage points over the strongest prior guard. Code, models, and evaluation artifacts will be available at https://yunhao-feng.github.io/HazardAuditor/.
Read the full abstract ▾
All authors Yunhao Feng, Ruixiao Lin, Ming Wen, Yanming Guo, Xingjun Ma, Yutao Wu, Xinhao Deng, Shouling Ji
Artificial Intelligence · Cross-listed from cs.CV
AdaVSkip: Adaptive Visual Token Skipping Across Layers For Efficient MLLMs Inference
Yuyao Sun, Tao Deng, Shuang Li et al. (+1 more)
Multimodal large
language models (MLLMs) require substantial computation to process numerous visual
tokens across all
transformer layers. Most methods for efficient MLLM inference exploit horizontal redundancy by compressing visual
tokens. Beyond
token reduction, recent studies exploit vertical redundancy through early exit or fixed-layer skipping. However, we find that the extent and distribution of this redundancy vary across inputs and differ between self-
attention and MLP modules. Motivated by these observations, we propose AdaVSkip, which equips each layer with two lightweight routers that independently determine whether visual
tokens pass through by or skip the self-
attention and MLP modules. These decisions collectively define an input-specific visual-computation path, but their discrete and non-differentiable nature makes learning effective paths challenging. To address this challenge, we develop a progressive two-stage training framework that updates only the routers while keeping the backbone frozen. Stage I establishes an initial routing policy through supervised training with input-specific targets derived from module-wise necessity scores. To further align the routing policy with task performance, Stage II uses reinforcement learning to optimize routing decisions with direct feedback from generated answers. It combines an answer correctness reward with a skip-consistency reward that discourages excessive retention of visual-
token computation. Across three MLLM backbones, AdaVSkip maintains strong task performance with substantially less computation. On LLaVA-NeXT-7B, AdaVSkip reduces FLOPs by 53.2% while preserving the original model's average performance. Combining it with visual
token compression increases this reduction to 91.2%, while retaining 97.2% of the original performance on average.
Read the full abstract ▾
All authors Yuyao Sun, Tao Deng, Shuang Li, Deqing Wang
Artificial Intelligence
Medical Knowledge Simplification for Patients in the Era of LLMs: A Case Study on Diabetes
Pallika Kafle, Yipeng Zhou, Guanfeng Liu et al. (+2 more)
Complex medical information is often difficult for patients to understand, making effective medical knowledge simplification essential for improving patient comprehension, informed decision-making, and health outcomes. Recent advances in large
language models (LLMs) provide a promising approach for simplifying complex medical information into patient-friendly
language; however, their effectiveness in real-world patient education remains insufficiently explored through human evaluation. To investigate their practical effectiveness, this paper presents a case study on diabetes knowledge simplification through the implementation and evaluation of MediClear, an LLM-based medical knowledge simplification system enhanced with
Retrieval-Augmented Generation (RAG). Public diabetes-related articles from Diabetes Australia, WHO, American Diabetes Association (ADA), NIDDK, and AIHW are indexed in the RAG knowledge base to retrieve clinically grounded information, which is then simplified by the LLM into accessible patient explanations. We evaluate the generated responses using standard readability metrics, including the Flesch-Kincaid Grade Level (FKGL), and conduct a human study involving 10 participants. Results show that MediClear consistently reduces the reading level of generated responses to the recommended patient literacy range while achieving high user satisfaction and willingness for future use. This case study demonstrates the potential of LLMs to improve the accessibility of medical knowledge for patient education.
Read the full abstract ▾
All authors Pallika Kafle, Yipeng Zhou, Guanfeng Liu, Quan Z. Sheng, Cheng-Hsin Hsu
Artificial Intelligence · Cross-listed from cs.CL
MoME: Mixture-of-Memory Embeddings for Context-Aware Sparse Lookup
Muchen Li, Leonid Sigal and Renjie Liao
Scaling large
language models efficiently has motivated sparse capacity mechanisms such as Mixture-of-Experts and, more recently, conditional
memory:
token-indexed
embedding tables that augment the backbone with cheap parametric lookups. Existing
memory-
embedding methods retrieve via a deterministic function of the surface form, which collapses different contextual senses of the same
token (e.g., python the
language vs. the animal) into a single fixed entry. We introduce Mixture of
Memory Embeddings (MoME), a context-aware
memory mechanism that replaces each
token's single
memory row with a mixture of M slots and uses a learned gate over the hidden state to choose which slots to read at each position. In controlled pretraining experiments across nanochat, Llama-3/MobileLLM, and Qwen3 backbones, MoME improves over Value
Embedding, Bigram, and STEM baselines in iso-parameter and iso-training-FLOP settings, shows a more promising
memory-size scaling trend at sub-billion scale, and remains efficient in training and inference. Qualitative routing analyses on polysemous
tokens further suggest that the learned mixture exhibits a degree of semantic interpretability, dispatching the same surface
token to distinct
memory slots under different senses.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.LG
Refinement-Based Flow Policy Optimization
Bumgeun Park, Hyukjun Yang and Donghwan Lee
Flow-based policies offer an expressive representation for online reinforcement learning, but conventional flow matching requires samples drawn from the distribution to be modeled. This poses a challenge when the desired action distribution is defined only implicitly by a Q-function, since directly sampling actions from the resulting distribution is generally intractable. We propose Refinement-Based Flow Policy Optimization (RFPO), a novel framework for training a flow policy in online reinforcement learning by alternating between Q-guided sample refinement and self-target flow matching. RFPO first generates actions from Gaussian noise using the current flow policy and then uses a finite-step
stochastic refinement procedure to move them toward an energy-based distribution induced by the Q-function. Each refined action is then paired with its corresponding initial noise sample and used as a fixed target for flow-matching training. By repeatedly refining its own outputs and learning from the resulting targets, RFPO incorporates Q-guidance into the policy without requiring direct samples from the target distribution, while retaining the capacity to represent multiple action modes. We further provide a theoretical analysis of the distributional dynamics induced by RFPO. Across six continuous-control tasks, RFPO matches or outperforms a standard Gaussian-policy baseline on almost every task. Experiments on six synthetic two-dimensional target distributions with diverse geometries demonstrate that RFPO captures complex
multimodal structure without mode collapse.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.SE
DepthBenchCAD: When Does Deeper Auditing Yield More Reliable Conclusions?
Hongye Yang, Zhihao Xie, Shengjun Xiong et al. (+1 more)
Generative CAD models are expected to remain behaviorally correct after parameter edits, so increasing the number of edit checks is often treated as a direct route to more reliable evaluation. Under a fixed budget, however,
auditing each program more thoroughly reduces the number of tasks and independent generations that can be evaluated, which can ultimately make model-level estimates less accurate. We study this phenomenon and the conditions under which it arises. We decompose behavioral evaluation into three evidence levels: task templates,
stochastic generations, and within-program edits. We define an average failure risk that is invariant to audit depth, and combine three-level variance with measured execution costs to analyze the tradeoff between deeper edit
auditing and broader independent coverage. Experiments across two CAD environments and five generation systems show that the value of deeper
auditing depends on where evaluation uncertainty originates. When template heterogeneity or generation stochasticity dominates, additional edit checks can increase total estimation error; when within-program state variation is large and generation is expensive, deeper
auditing is more valuable. Variance and cost estimates from calibration predict the direction of this change and provide a diagnostic basis for allocating evidence on held-out tasks. These results show that the thoroughness of program inspection can diverge from the reliability of model evaluation, and they help determine whether the next unit of budget should be spent on a new task, a new generation, or additional edit checks.
Read the full abstract ▾
All authors Hongye Yang, Zhihao Xie, Shengjun Xiong, Boxiao Huang
Artificial Intelligence · Cross-listed from cs.RO
Legislating World-Model-Based Planning with Legal Reasoning
Dylan Waldner, Yiannis Kantaros, Guido Governatori et al. (+2 more)
As robotic systems grow more general, legal norms are needed to integrate them into society. This paper extends the isomorphism problem of aligning legal source texts with their encodings, and measures two key challenges to
robot normative control: (1) the grounding isomorphism gap, where perception error grounds false atoms for legal
reasoning, and (2) the ontological isomorphism gap, where one legal conclusion admits many faithful translations into planning constraints. The paper introduces a legal planning stack that employs Defeasible Deontic Logic (DDL) to constrain a motion planner. The stack leverages learned world models to plan and to provide legal context, enabling ex ante governance that intervenes before an illegal action is executed. It was deployed on a simulated
robot arm pushing a cube across a 3x3 grid. The findings were (1) the legislated
agent abided substantially more often than the non-legislated one, and modeling perception uncertainty lifted abidance even further, (2) the legal
reasoning ran efficiently at runtime and its verdicts were auditable, and (3) the stack adapted to exogenous signals and endogenous rule changes. Both gaps were measured: (4) world model and probe error corrupted the factual input for the DDL reasoner, and (5) a single law admitted several faithful metric interpretations yielding drastically different abidance. Thus, ex ante legislation functions as intended, and closing these gaps with a standardized mapping from the law to runtime constraints and improved fact grounding from perception will yield robust laws that align
robot behavior with society's norms. Project page: https://dylanwaldner-cail.github.io/Legislated-Planner/.
Read the full abstract ▾
All authors Dylan Waldner, Yiannis Kantaros, Guido Governatori, Risto Miikkulainen, Amir Banifatemi
Artificial Intelligence · Cross-listed from physics.flu-dyn
Physics Informed Neural Network model for the dynamical study of Abdominal Aortic Aneurysm
Adrián Robles Arques, Martín Ruiz Fernandez, Javier Sanchis et al. (+2 more)
We present the development and application of a three-dimensional Physics-Informed Neural Network (PINN) framework for the investigation of haemodynamic behaviour in the human aorta. The model incorporates a time-resolved simulation of pulsatile blood flow over a two-minute interval, enabling the extraction of pressure and velocity fields with high temporal fidelity. The mechanical stress exerted on the aortic wall was quantified through Laplace's law, with temporal averaging applied to derive representative stress distributions. This approach circumvents the computational overhead associated with conventional computational fluid dynamics (CFD) methods by eliminating mesh generation and exploiting the automatic differentiation capabilities inherent to neural networks. The proposed methodology demonstrates that PINNs can serve as an efficient and accurate alternative for modelling complex vascular flow phenomena, offering significant advantages in scalability and computational cost reduction while maintaining physical consistency.
Read the full abstract ▾
All authors Adrián Robles Arques, Martín Ruiz Fernandez, Javier Sanchis, Miguel A. Teruel, Juan Trujillo
Artificial Intelligence · Cross-listed from cs.CV
ChatGPT Images 2.5 in the Wild: A Launch-Period Dataset and Detector Evaluation
Dennis Ng, Xingyu Shen, Ankit Raj et al. (+6 more)
An image tool can change its underlying generator while retaining its public name, making version attribution from online posts ambiguous. We study this problem after the ChatGPT Images 2.5 launch. Our frozen collection contains 3,478 images from 2,440 posts across 8 sources. Recorded posting times fall within the first 51.1 hours after the announcement. It records three attribution tiers and retains standalone images after image-form filtering and targeted review. Caption claims and host records provide admission evidence, not independently verified generator identity. The observed content profile depends on the source mixture: NightCafe supplies 39.0% of images but 77.0% of CLIP-assigned fantasy scenes. We then evaluate six frozen detectors at thresholds calibrated to a 5% flag rate on reference photographs. Collection flag rates range from 3.7 to 56.4%, falling 42-81 percentage points below GenImage recall. Held-out artwork false-positive rates range from 1.5 to 96.5%, so a higher collection flag rate does not by itself establish better detection. An exploratory X-only comparison with our April collection finds a higher September flag rate for Effort, and a suggestive difference for DoU, under fixed-threshold post-clustered bootstrap intervals. Attribution, content and processing differences prevent a causal interpretation of these contrasts. The collection supports analysis of reported model use during a product transition, with source and attribution evidence retained for interpretation. The collection is released at https://scam.ai/research.
Read the full abstract ▾
All authors Dennis Ng, Xingyu Shen, Ankit Raj, Kidus Zewde, Tommy Duong, Yuchen Zhou, Yuxin Zhang, Neo Tiangratanakul, Simiao Ren
Artificial Intelligence
OpenAI4S: Code as Action, Science as Sessions
Gongbo Zhang, Hao Li, Yu Wang et al. (+15 more)
AI co-scientists could accelerate computational research, but over a long-running study the workflow also has to stay inspectable, resumable and reproducible, which requires persistent computational state and provenance. Here we present OpenAI4S, an open-source scientific research
agent built around the principle of Code as Action, Science as Sessions. OpenAI4S combines a persistent computing runtime with research-session management: orchestration is handled through structured tool calls, while scientific actions are represented as complete code
cells executed in persistent Python and R kernels. An append-only Action Ledger, per-
cell execution records, versioned artifacts, environment records, and workspace checkpoints preserve how results were produced and support session recovery, branching, and extension. Configurable sandboxing, permission controls, and code and trajectory screening provide complementary safeguards. We evaluate OpenAI4S on 36 research scenarios spanning retrosynthesis, molecular dynamics,
protein binder design,
protein mutation, catalyst screening, and mineral spectroscopy, measuring scientific task accuracy, workflow completeness, and reproducibility of the resulting repositories. OpenAI4S achieves an overall score of 7.83, compared with 5.7–6.4 for a general-purpose coding harness evaluated with three frontier models, with the largest gains on long-horizon and computation-intensive workflows. These results suggest that integrating persistent execution with session-level provenance can improve the reliability of AI-assisted scientific workflows. Environment specification and full rerunnability remain weak for every evaluated system, ours included, so reproducibility is still an open problem for scientific
agents. The system is available under the MIT license at \href{https://github.com/PKU-YuanGroup/OpenAI4S}{github.com/PKU-YuanGroup/OpenAI4S}.
Read the full abstract ▾
All authors Gongbo Zhang, Hao Li, Yu Wang, Mujie Lin, Liuzhenghao Lv, Yicheng Mao, Yimi Wang, Jun Zhu, Minhan Tang, Zhengxiang Jiang, Yusong Wang, Jiayu Yao, Kunpeng Ning, Dawei Pang, Yonghong Tian, OpenAI4S Community, Yuyang Liu, Li Yuan
Artificial Intelligence · Cross-listed from cs.IR
Generate to Explore, Select to Exploit: Aligning LLM-based Headline Generation with Personalized Recommendation
Yi Chen, Rufeng Cheng, Qiang Xie et al. (+1 more)
In industrial recommendation feeds, presenting a static headline for an item often fails to satisfy the diverse,
multimodal interests of the user population, particularly suppressing the needs of long-tail audiences. While Large
Language Models (LLMs) have been integrated into recommendation for content understanding or ranking, directly optimizing them to output a single best headline typically leads to mode collapse—converging to generic patterns that satisfy average tastes but miss specific latent intents. To bridge this gap, we introduce GESE (Generate to Explore, Select to Exploit), a framework operating at the system's presentation layer that decouples personalization into generative exploration and selective exploitation. First, we treat the LLM as a probabilistic explorer, utilizing Group Sequence Policy Optimization (GSPO) with a hierarchical reward mechanism to generate a candidate set that maximizes the semantic coverage of potential user interests. Subsequently, a lightweight, real-time feedback-aware selector acts as the exploiter, identifying the optimal realization from the candidate pool based on instant contextual signals. Extensive deployment on a commercial platform with over 100 million daily active users demonstrates that GESE significantly outperforms state-of-the-art baselines, achieving a 2.57% lift in CTR and 0.87% in dwell time. These results validate that decoupling diversity-oriented generation from precision-oriented selection offers a robust blueprint for aligning generative AI with dynamic user utility.
Read the full abstract ▾
All authors Yi Chen, Rufeng Cheng, Qiang Xie, Tao Li
Artificial Intelligence
ER-EDF: A Psychology-Grounded Emotion Regulation Framework for Speech Empathetic Dialogue Generation in Large Audio-Language Models
Hongyu Jin, Wenda Zhang, Runqiu Fei et al. (+3 more)
In human conversation, empathy involves not only recognizing a partner's
emotion but also regulating it, listeners identify the emotional state, select a suitable regulation strategy, and implement it in their reply. Current large audio-
language models (LALMs) largely map perceived
emotion directly onto a response, skipping this regulation process. As a result, their responses tend to mirror the user's affect or offer generic sympathy. We propose ER-EDF, a psychology-grounded framework for empathetic dialogue that follows the three stages of Gross's Extended Process Model of
emotion regulation, identifying the user's
emotion from
speech, selecting a regulation strategy matched to that
emotion, and implementing the strategy through a LALM. The framework is model-agnostic and integrates seamlessly into existing LALMs. We further construct a spoken empathetic dialogue
dataset and introduce empathy-aware evaluation metrics beyond lexical matching. Experiments across five LALMs and two
datasets show that ER-EDF consistently improves empathetic response quality in both automatic and human evaluations. These results suggest that modeling how humans regulate others'
emotions, rather than only how they perceive them, is key to empathetic spoken dialogue, paving the way for psychologically grounded empathetic AI.
Read the full abstract ▾
All authors Hongyu Jin, Wenda Zhang, Runqiu Fei, Gongping Huang, Mike Conway, Ting Dang
Artificial Intelligence · Cross-listed from cs.LG
Beyond Numerical Time Series: A Unified Benchmark for Multimodal Forecasting with Heterogeneous Context
Peng Chen, Zhihao Zhuang, Hongzhou Chen et al. (+6 more)
Most time series
forecasting benchmarks remain numerical-centric and provide limited support for evaluating contextual information that shapes real-world temporal dynamics. Existing
multimodal benchmarks also suffer from limited data and context coverage, fragmented evaluation settings, and overreliance on aggregate evaluation. In this paper, we propose MUSE-Bench, a unified
benchmark for
multimodal time series
forecasting with heterogeneous context. It comprises fourteen
datasets across eight domains and six types of context: metadata, events, holidays, news, images, and numerical
covariates. We evaluate diverse
forecasting paradigms, including statistical, data-specific, foundation,
multimodal, and general-purpose LLM
forecasting methods under shared non-overlapping forecast windows, common target observations, and consistent point and probabilistic metrics. Extensive experiments yield three main findings. First, numerical time series foundation models dominate the overall ranking, while Aurora, the evaluated
multimodal foundation model, trails the leading numerical TSFMs but outperforms all evaluated data-specific models. Second, ablations show that external context improves the four evaluated context-aware models, whereas incorrect or temporally misaligned context degrades performance. Third, general-purpose LLMs perform poorly as direct forecasters, and LLM-guided refinement does not yield consistent improvements. MUSE-Bench enables systematic evaluation of how
forecasting models utilize context and provides a foundation for future
multimodal forecasting research.
Read the full abstract ▾
All authors Peng Chen, Zhihao Zhuang, Hongzhou Chen, Junhao Huang, Aiping Yang, Mengsen Wu, Yiding Liu, Xilin Dai, Zewei Dong
Artificial Intelligence · Cross-listed from cs.CL
Translating the Translator: Decomposing the Cost of English-Forced Inter-Agent Communication
Kushagra Agrawal, Yuming Feng and Man-Fai Leung
Multi-
agent LLM architectures, such as LangChain and AutoGen, largely assume English as the lingua franca for internal inter-
agent communication, even when the end-user task is non-English. We fill this gap by evaluating a two-
agent extraction-answer core, with an additional back-translation
agent in the English-forced condition, across four typologically diverse
languages (Hindi, Chinese, Spanish, Arabic; n = 300 per
language) using the Aya-23-8B model. We compare a native-
language pipeline to an English-forced one (which incorporates a final back-translation step from English to the user's
language). We discover a statistically significant English-Forcing Tax (surviving a strict Bonferroni correction) that isolates the cost of English routing from general multi-
agent orchestration overhead. Forcing inter-
agent communication through English reduces Exact Match accuracy by 13.0 percentage points (Spanish) up to 30.6 percentage points (Hindi) compared to native-
language multi-
agent execution. Using chrF scores as a diagnostic measure of English-reference lexical overlap, we find that lower overlap is strongly associated with pipeline failure, consistent with translation loss being an important contributor to the observed performance drop. These findings suggest a compelling case for native-
language routing in
agent frameworks when the source and target
languages are typologically distant, reducing a compounding translation tax.
Read the full abstract ▾
Artificial Intelligence
Enabling Creative Exploration for Vibe Design Agents
Yifan Zhang, Nghi D. Q. Bui, Georgios Evangelopoulos et al. (+1 more)
Vibe design
agents translate natural-
language requests into interfaces, yet repeated generations often converge on similar visual directions. We present an inference architecture that makes design exploration an explicit stage before implementation. A pre-pass proposes structured design specifications with typicality weights, an external policy samples a direction, and the generator implements it under fixed decoding settings. Candidate elicitation draws on Verbalized Sampling. Applied to UI themes and visual-asset
prompts, this architecture provides direct control over aesthetic exploration without changing code-generation temperature. Across 168
benchmark prompts, theme sampling expands observed selection coverage from one option to nearly three and increases rendered visual variation. For visual-asset
prompts, the strongest observed LLM-judge win/loss ratio is 1.32, with only a small change in whole-screen similarity. Localized aesthetic choices can thus influence judged preference even when the overall screen remains similar. We further evaluate joint theme and asset sampling in a randomized online study conducted over nearly two months, covering 3.39 million tasks and 87,416 labeled conversations. These studies show how explicit design specifications support exploration of alternative UI concepts and help us evaluate their effects on rendered interfaces, judged preference, and user interaction.
Read the full abstract ▾
All authors Yifan Zhang, Nghi D. Q. Bui, Georgios Evangelopoulos, Arnaud Benard
Artificial Intelligence · Cross-listed from cs.LG
Branched Optimal Transport Amortization
Semyon Semenov, Viktor Kovalchuk, Meir Roketlishvili et al. (+4 more)
Methods of Branched Optimal Transport (BOT) mimic the economy and efficiency of natural tree-like structures, such as those found in rivers and biological systems. These methods are widely applicable for designing efficient networks in society, from river basins and blood vessels to mail and gas distribution systems. However, they remain understudied in the context of designing deep generative models, particularly at a large scale. Standard continuous-time generative models, such as the flow matching approach, fail to capture the inherent hierarchical and branching patterns present in real-world data. Current models provide no mechanism for flows to merge or share pathways to minimize total transport cost. Inspired by the "economy of scale" principle in BOT, we introduce a novel, scalable branched flow-matching
algorithm designed to solve the branched optimal transport problem in high dimensions. Our method adapts the Benamou-Brenier continuous-time optimal transport formulation to learn branched generative flows. These flows allow probability mass to aggregate along common pathways before branching out to diverse targets. Parametrized by neural networks, our method effectively learns complex branched generative processes. We demonstrate its effectiveness on challenging high-dimensional tasks in biology and image generation.
Read the full abstract ▾
All authors Semyon Semenov, Viktor Kovalchuk, Meir Roketlishvili, Albert Baichorov, Fakhri Karray, Martin Takac, Arip Asadulaev
Artificial Intelligence · Cross-listed from cs.SD
Rethinking Procedural Audio Pre-training: Source Scaling and Objective Adaptation
Jiajun Peng, Fengrui Liu, Xinyu Liu et al. (+1 more)
Procedural audio has emerged as a viable source for transferable audio representation learning, but its design principles remain unclear.We revisit two questions: how a procedural source should be scaled, and whether training choices developed on natural audio should transfer unchanged to procedural data.Using a controlled source, we separate scale into formula-class coverage C and within-class rendering diversity I.Experiments with FDSL and AudioMAE show that these two forms of scale provide different benefits and depend on the learning formulation and downstream task. A matched AudioMAE study further shows that procedural audio favors low mask ratios (10%–25%), whereas AudioSet-28K favors 50%–75%. Shared-codebook analysis reveals lower patch diversity and stronger temporal predictability in procedural audio. These results motivate source-aware procedural pre-training, where source scaling and learning configuration are considered jointly.Code is available at https://github.com/Cross-Innovation-Lab/Formula-Bank.
Read the full abstract ▾
All authors Jiajun Peng, Fengrui Liu, Xinyu Liu, Feng Liu
Artificial Intelligence · Cross-listed from cs.CL
Salesforce Koa: An Enterprise Language Model for Agentic Tool Use
Zixiang Chen, Sufeng Niu, Yingchi Liu et al. (+23 more)
We present Salesforce Koa, an enterprise
language model built by post-training the open-weight Nemotron-3-Super-120B foundation model with reinforcement learning using Group Relative Policy Optimization (GRPO), and deployed in FP8 for production. Koa is trained only on public and synthetically generated data, and specialized for the agentic tool use that enterprise workflows demand: routing a request to the correct action, invoking the right tool with valid arguments, and completing multi-turn business tasks. The distinctive component of our pipeline is specification-driven task construction: declarative
Agent Script specifications are expanded into persona-conditioned multi-turn environments whose rewards are grounded in successful tool use. Applied to enterprise CRM specifications, the same pipeline produces the in-domain training distribution on which Koa is specialized. On CRMAgentBench and the human-labeled production tool-calling set, Koa outperforms both its untuned open-weight base and GPT-4.1 and is competitive with the strongest frontier models. It reaches 87% Task Success Rate on CRMAgentBench (vs. GPT-4.1 at 82% and the base at 79%), is at or near the top of every metric on the human-labeled portion of an internal production
benchmark, and preserves the base model's general capability on public
benchmarks (Tau2Bench, BFCL). A controlled comparison with architecture and RL recipe held fixed shows that the additional in-domain RL stage improves argument accuracy and full tool-call success on the human-labeled enterprise
benchmark.
Read the full abstract ▾
All authors Zixiang Chen, Sufeng Niu, Yingchi Liu, Wenting Zhao, Akshara Prabhakar, Shubham Mehrotra, Bin Bi, Zhujun Lan, Katherine Tan, Mohammad Ramezanali, Tulika Manoj Awalgaonkar, Monojit Banerjee, Jielin Qiu, Shiva Kumar Pentyala, Zhepeng Cen, Anupam Tripathi, Ali Ziaei, Regunathan Radhakrishnan, Darvish Lee Shadravan, Shelby Heinecke, Sitaram Asur, Jayesh Govindarajan, Silvio Savarese, James Zhu, Phil Mui, Huan Wang
Artificial Intelligence
BusMA: A Bus Communication Substrate for Multi-Agent Systems
Yanwen Peng, Delvin Ce Zhang, Xi Wang et al. (+1 more)
Multi-
Agent (MA) systems are effective at solving complex tasks that demand planning, tool use, and the synthesis of evidence from multiple sources. Existing systems typically adopt Hierarchical Manager-Worker (HMW) or Router-based Message Passing (RMP) structures as their communication protocol. However, these designs restrict
agent autonomy: Worker
agents cannot directly consult specific "peers", and misrouted messages can propagate errors. Inspired by bus architectures in computer systems, we propose BusMA, a communication framework that allows any
agent to address other
agents through a shared channel, i.e., the Bus. It consists of
agent registration, message routing, and shared
memory management components. Worker
agents, each equipped with tools, have their own local
memory and can reason, act (tool usage), and communicate by posting shared messages with specific intents. We introduce four intents: discussion, challenge, guidance, and request for explanation, which support fine-grained communication among
agents. A Chair
agent monitors the shared
memory to coordinate interactions and facilitate convergence among Workers. To evaluate the effectiveness of BusMA, we conduct extensive experiments with two frontier LLMs across 13 tasks spanning visual
reasoning, mathematical
reasoning, and knowledge
retrieval. The results demonstrate that BusMA consistently outperforms state-of-the-art HMW and RMP methods.
Read the full abstract ▾
All authors Yanwen Peng, Delvin Ce Zhang, Xi Wang, Nikolaos Aletras
Artificial Intelligence · Cross-listed from cs.LG
Not All Prompts Are Equal: Exploration-Guided Prompt Scaffolding for Multimodal Reinforcement Post-Training
Yuanhao Yue, Qianli Ma, Chengyu Wang et al. (+3 more)
Training
prompts in online reinforcement learning (RL) differ substantially in how informative they are for the current policy: some are already saturated while others are too difficult to yield reliable learning signals, yet both receive equal rollout budget under standard training. We propose an exploration-guided
prompt scaffolding framework that adapts the training
prompt distribution dynamically throughout RL post-training of
multimodal large
language models (MLLMs). Central to our approach is the $\textit{Exploration Potential Score} (EPS)$, a lightweight rollout-based proxy for
prompt utility derived from KL-regularized policy improvement theory, computable directly from on-policy rollout statistics without additional overhead. Rather than discarding low-utility
prompts, we use a teacher model to generate scaffolded rewrites that preserve the original task intent while making subsequent training more informative, reframing teacher supervision as training-data refinement rather than output imitation. Integrated with GRPO on Geo3K and MMK12, our method consistently outperforms the baseline on both in-domain and out-of-distribution
benchmarks, achieving up to 9.7% relative improvement in-domain and gains of 11.5% on MathVision and 11.1% on MMMU-Pro.
Read the full abstract ▾
All authors Yuanhao Yue, Qianli Ma, Chengyu Wang, Haoting Wang, Lei Shen, Jun Huang
Artificial Intelligence · Cross-listed from cs.LG
Ensemble Complexity in Photovoltaic Forecasting
Sun Ze, Zhou Liguo, Xu Yuqing et al. (+2 more)
An ensemble can improve photovoltaic forecasts while adding components that contribute little or increase computation. We assess these effects through matched comparisons and ablations of a fixed heterogeneous predictor bank. Hourly experiments use GEFCom2014 and three additional public
datasets, with chronological partitions and three seeds. Under retrospective ERA5 assistance, static fusion reduces scaled mean absolute error against matched boosting by 1.11%, 4.41%, and 1.63% on PVDAQ, OPSD, and Ausgrid; only OPSD remains supported after multiple-comparison correction. Weather gating offers no consistent incremental benefit. Exploratory member removals show group-level dependence alongside individual redundancy. A separate, previously inspected fifteen-minute case replaces one neural member with a tree predictor: normalized error falls by 1.72%, but measured inference is slower. These findings support component-wise evaluation with explicit limits on weather availability and test-set reuse.
Read the full abstract ▾
All authors Sun Ze, Zhou Liguo, Xu Yuqing, Yu Lei, Jiang Mingming
Artificial Intelligence · Cross-listed from cs.HC
Personalizing Personal Health Interfaces: Co-Design with Generative AI
Karthik S. Bhat, Vidhi Shah, Vedika Agnihotri et al. (+2 more)
Personal health interfaces present wellbeing data through standardized dashboards that rarely fit how people interpret or act on it. Personalizing them to what people would like to see for themselves often requires design and technical expertise, a barrier that generative AI may potentially lower. Therefore, we ask what designs emerge and how it enables and constrains the design process. We conducted a co-design study where 14 participants redesigned Google and Apple Health interfaces using Figma Make. Participants reimagined interfaces that supported personal context, future planning, and interactive experiences, yet conversational AI designs converged around chat-window conventions. AI helped materialize loosely articulated ideas, but model defaults and generation latency shaped iteration. The process more readily operationalized interpretability and accountability than
privacy, trust, and emotional safety. Generative co-design let participants create interfaces directly, blurring the boundary between intentions and model defaults. We discuss implications for preserving agency and flexible user-directed interfaces.
Read the full abstract ▾
All authors Karthik S. Bhat, Vidhi Shah, Vedika Agnihotri, Dong Whi Yoo, Koustuv Saha
Artificial Intelligence · Cross-listed from cs.CL
Mirror, Mirror on the Wall: Prompt Echoing in Small Instruct Language Models
Inez Okulska, Bartosz Naskręcki, Jan Piotrowski et al. (+1 more)
Prompt echoing is a recognized failure mode of instruct
language models, in which a model instead of generating a response, mirrors the provided
prompt, even though it did not receive a specific instruction to do so. Is this phenomenon a sign of the model leaking the content of its training
dataset, or is it rather caused by a misaligned behavior of the internal induction/copying mechanisms? We investigate
prompt echoing small
language models from different families (Gemma, Llama, Qwen, SmolLM and OLMo) and show that echoing
prompts are likely to have partial overlap with the training
dataset but the phenomenon is primarily driven by the model's induction heads.
Read the full abstract ▾
All authors Inez Okulska, Bartosz Naskręcki, Jan Piotrowski, Tomasz Steifer
Artificial Intelligence · Cross-listed from cs.CR
SpliTEE: Fast and Private LLM Inference by Coupling GPU-Assisted Trusted Execution Environments with Differential Privacy
Shashie Dilhara Batan Arachchige, Robin Carpentier, Hassan Jameel Asghar et al. (+1 more)
User
prompts provided to large
language models (LLMs) may contain private information. One way to protect them is to execute the LLM inside a trusted execution environment (TEE). However, this results in slow inference times as current TEEs are significantly slower than GPUs for LLM inference. To circumvent this, Tramèr and Boneh (2019) proposed Slalom which splits neural network inference between a TEE and an untrusted GPU. They encrypt inputs to computations outsourced to the GPU. In this paper, we extend this split-inference architecture to LLM inference and instead protect intermediate inputs using differential
privacy (DP). We first demonstrate that masking intermediate representations is necessary by showing an 80% accuracy on a
prompt-reconstruction attack from these representations. Our main contribution is a global sensitivity analysis of key functions in LLM inference, which bounds the required scale of DP noise. Unlike encryption, DP avoids quantization, allowing the LLM to remain in the floating-point domain. We also derive an upper bound on the floating-point error from masking and subsequent noise cancellation as a function of the
privacy parameter epsilon, keeping the same quality of the LLM response. We implement our architecture using the Intel TDX TEE and two LLMs: Llama-3.2-3B and Qwen3-4B. Our split execution is nearly twice as fast as fully TDX-based inference. Moreover, it is at most 43% faster than Slalom while achieving higher accuracy. Finally, we demonstrate that
prompt reconstruction, even with knowledge of the DP mechanism, cannot recover more information than is contained in an unrelated
prompt.
Read the full abstract ▾
All authors Shashie Dilhara Batan Arachchige, Robin Carpentier, Hassan Jameel Asghar, Dali Kaafar
Artificial Intelligence
The average-farmer illusion in language-model simulations of agricultural decisions
Zhanliang Zhu, Ziwei Li, Yuchen Liu et al. (+5 more)
Language-model
agents are increasingly used as synthetic people in surveys and social simulations, yet their apparent realism is often judged from population averages or distributional similarity. We tested what such evidence actually establishes by comparing Claude, Codex and Kimi under four prespecified
prompt designs with matched farmer decisions from China and four African countries. Some configurations reproduced observed means and adoption rates. However, their person-level predictions were weak; their decisions clustered around typical values and policy-relevant extremes were largely missing. Most strikingly, a simple generator fitted only to the observed marginal dis- tribution, and given no information about any farmer, achieved greater distributional similarity than every
language-model configuration.
Prompt additions produced conditional gains rather than uni- versal improvement: results varied with model, outcome, population and validation target. We call this the average-farmer illusion: a synthetic population can look realistic while failing to repro- duce who does what or how behaviour varies. We provide a claim-matched validation framework and reusable modular
prompts that turn
prompt construction into an auditable experimental process. Population-level resemblance should therefore be treated as the start of validation, not as evidence of individual simulation.
Read the full abstract ▾
All authors Zhanliang Zhu, Ziwei Li, Yuchen Liu, Liujun Zhu, Ruiqi Wu, Tongqing Shen, Junliang Jin, Jianyun Zhang
Artificial Intelligence
Horizon-specific Expert Fusion for Photovoltaic Power Forecasting
Xu Yuqing, Zhou Liguo, Sun Ze et al. (+2 more)
Short-term photovoltaic power
forecasting requires models to represent regular
solar cycles and weather-driven fluctuations whose importance changes with the forecast horizon. This study develops a hierarchical ensemble that combines temporal neural models, historical analogs, state climatology, and gradient-boosted trees.
Solar geometry and numerical weather forecasts describe the expected generation conditions, while horizon-specific convex weights combine complementary predictions. A separate calibration step uses available historical forecast errors to account for recent bias. The framework is evaluated on public PVDAQ data at 15–240-minute horizons and on three GEFCom2014
solar zones at hourly horizons up to four hours. On PVDAQ, the ensemble achieves a daylight capacity-normalized mean absolute error of 4.315%, reducing error by 4.11% relative to full-feature LightGBM and by 6.03% relative to fine-tuned Chronos-2 under identical calibration. Expert-removal experiments identify redundancy within the ensemble. Across three training seeds on GEFCom2014, learned fusion improves upon equal weighting but performs comparably to LightGBM. The results support horizon-specific combination as a useful
forecasting strategy while showing that its advantage over strong individual models depends on the
dataset and evaluation period.
Read the full abstract ▾
All authors Xu Yuqing, Zhou Liguo, Sun Ze, Yu Lei, Jiang Mingming
Artificial Intelligence · Cross-listed from cs.DC
Validating Hybrid-State Cache Recovery for GLM-5.3-Flash with vLLM and LMCache
Frank Li
External cache transfers can succeed while a hybrid
language model resumes from an inconsistent state. We examine the full 45-layer GLM-5.3-Flash model, using the RedHatAI/ GLM-5.3-Flash-NVFP4 quantized checkpoint with vLLM and LMCache under four-way tensor parallelism. A complete-hit recovery mismatch restored state for the full
prompt while the scheduler credited one fewer
token. We aligned recovery through strict-prefix lookup and established a numerical comparison using shared computation corrections, matched checkpoint scheduling, and fixed per-rank kernel configurations. In a nine-length serial workload, agreement with the modified recomputation control improved from 34/36 to 36/36 generations, each containing 64
token IDs. A separate instrumented run passed recorded transfer-page, effective-tail, and delayed-save checks. Three additional synthetic templates passed 72 paired 256-
token continuations across two fresh-container runs. A subsequent serial performance study preserved output equality across 120 requests; among the measured trials, CPU reload reduced time to first
token by 46-64% and total request time by 1.9-7.0% relative to modified cold recomputation. The contribution is an experimentally validated integration repair applying an existing checkpoint-
alignment principle. The evidence is confined to one model revision and controlled configuration; it does not establish general determinism, task-quality equivalence, concurrent-serving gains, or capacity beyond GPU
memory.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.LG
Pick Your Poison: Learning to Select Poison Sets for Stronger LLM Backdoor Attacks
Aashiq Muhamed, Mona T. Diab, Virginia Smith et al. (+2 more)
Backdoor poisoning attacks add poisoned examples to otherwise-clean finetuning data, pairing a trigger with a target behavior that the model learns to produce when the trigger appears. Existing evaluations typically fix the number of poisoned examples and sample them at random from a candidate pool. We show that this can severely underestimate worst-case vulnerability: across three LLaMA-3-8B
backdoor settings, holding the model, clean data, and poison count fixed, attack success ranges from 3% to 80% depending only on which poison set is chosen. We formalize poison selection as oracle-budgeted set optimization and introduce SAILS (Set-level Audit-Informed Iterative Learned Selection), which learns a set scorer from a few hundred finetune-and-evaluate runs, ranks millions of candidate sets, and audits only a small shortlist. SAILS improves held-out attack success by 30 percentage points on average over the strongest influence baselines, transfers from small-scale to full-scale finetuning, and extends to code-generation, agentic, and API-only
backdoors.
Read the full abstract ▾
All authors Aashiq Muhamed, Mona T. Diab, Virginia Smith, Andrew Ilyas, Matthew Jagielski
Artificial Intelligence · Cross-listed from cs.CR
PIDS-Bench: Evaluating Prompt-Injection Detectors Under Over-Defense, Obfuscation, and Distribution Shift
Yusuf Khalid Shire and Sang-Chul Kim
Prompt-injection detectors are typically evaluated using aggregate F1 on in-distribution test data, which offers limited insight into behavior under distribution shift, particularly on the benign side of the decision boundary, where false positives impose direct operational cost yet are seldom measured. We present PIDS-Bench, a frozen multi-axis
benchmark that jointly evaluates attack detection and benign false-positive behavior at fixed thresholds, spanning in-distribution inputs, hard-benign
prompts that mimic injection structure without malicious intent, obfuscated attacks, and domain and structural distribution shifts. We evaluate seven detectors (learned baselines, external
prompt-injection classifiers, and broad-safety comparators) alongside a rule-based lower-bound reference. Multi-axis evaluation exposes a failure mode that aggregate F1 conceals. A detector exceeding F1 = 0.98 on the held-out split still misclassifies roughly one-third of an externally-sourced benign subset drawn from public corpora and restricted to security-adjacent content. Across a full threshold sweep and five training seeds, no internal detector reaches an operating point satisfying F1 >= 0.95 and hard-benign FPR <= 0.10 together on this stress distribution. Decomposing by provenance, we find that hard-negative augmentation nearly eliminates over-defense on curated stress inputs but leaves it substantially intact on externally-sourced
prompts, a pattern we term provenance-sensitive over-defense. The asymmetry holds across both fine-tuned architectures and does not diminish as the augmentation pool grows, with the externally-sourced FPR remaining far above the 0.10 target. Whether augmentation matched to the externally-sourced distribution would close this gap is untested; threshold calibration and curated-style augmentation alone do not.
Read the full abstract ▾