The paper introduces a retrieval‑augmented multi‑agent framework that automatically generates instance‑specific evaluation rubrics for medical language models. By retrieving authoritative medical evidence, decomposing it into atomic facts, and combining these with user interaction constraints, the system produces fine‑grained criteria that outperform GPT‑4o on HealthBench and LLMEval‑Med. The generated rubrics also guide response refinement, improving medical LLM output quality by 9.2%.
By Yinzhu Chen, Abdine Maiga, Hossein A. Rahmani, Emine Yilmaz
The paper presents a training‑free, agentic computer‑vision system that converts 2D structural framing plan PDFs into editable finite‑element model drafts. It uses a deterministic stage to extract geometric primitives, estimate scale, and recognize five entity classes via a drafting grammar, followed by an agentic stage that applies typed corrections and fail‑closed transactions. Evaluation on a 100‑plan benchmark shows high precision and recall across columns, beams, walls, braces, and openings, with scale estimates within 0.1% of reference.
By Mohammad Talebi-Kalaleh, Qipei Mei
The paper studies finite‑horizon minimum‑energy steering of inertial swarms under stochastic disturbances, focusing on mean‑field models with Cucker–Smale alignment or Morse attraction–repulsion interactions. It formulates the problem as a Schr"odinger bridge, deriving nonlinear, time‑symmetric optimality systems and proposing nested fixed‑point schemes for numerical solution. Numerical experiments demonstrate that the optimal corrective drift can either exploit or counteract the natural interaction forces, depending on their alignment with the steering objective.
By Asmaa Eldesoukey, Md Zulfiqur Haider, Italo Napolitano, Yongxin Chen, Abhishek Halder
EduAgentQG is a multi‑agent framework for generating personalized mathematics questions that explicitly controls diversity and aligns with educational objectives. It operates through a closed‑loop cycle of planning, writing, evaluation, refinement, and checking, using fine‑grained evaluation to ensure logical correctness, solvability, and alignment with knowledge concepts, difficulty, grade level, and core competencies. The authors built a benchmark of 10,273 questions across Grades 1‑9 and demonstrated that EduAgentQG outperforms existing methods in diversity, objective consistency, and win rate.
By Rui Jia, Min Zhang, Fengrui Liu, Bo Jiang, Kun Kuang, Zhongxiang Dai
The paper introduces RoMem, a temporal knowledge graph module that treats time as continuous phase rotation rather than discrete labels. RoMem uses a Semantic Speed Gate to assign volatility scores to relations, allowing evolving facts to rotate quickly while persistent facts remain stable, thereby preventing the need for deletion or costly LLM calls. The method achieves state‑of‑the‑art performance on ICEWS05‑15 and improves temporal reasoning in agentic memory benchmarks such as MultiTQ, LoCoMo, and FinTMMBench.
By Weixian Waylon Li, Jiaxin Zhang, Xianan Jim Yang, Tiejun Ma, Yiwen Guo
The paper introduces Gavel, a framework for evaluating large language models (LLMs) on long-context legal summarization tasks. Gavel includes a reference-based component (Gavel-Ref) with checklist, residual-fact, and writing-style checks, and a reference-free component (Gavel-Agent) that assesses factual coverage directly from source documents. Experiments on 12 frontier LLMs reveal that models tend to omit key information more than hallucinate, perform well on simple checklist items but struggle with rare, complex items, and their performance degrades with longer cases. Gavel-Agent cuts token usage by at least 36% compared to traditional methods while maintaining competitive accuracy, and it also generalizes effectively to the medical domain.
By Yao Dou, Benjamin Mamut, Wei Xu
The paper introduces AWM, a framework that treats the terminal working memory of long‑document VQA agents as an answerable evidence artifact. It proposes a memory‑only answerability diagnostic and incorporates this signal into the GRPO reward, giving higher advantage to trajectories whose final memory can answer the question alone. Experiments on MMLongBench‑Doc and LongDocURL show that AWM‑GRPO boosts final‑answer accuracy by up to 11.9 points and reduces the rate of correct answers that cannot be supported by memory alone.
By Dongzhuoran Zhou, Yuqicheng Zhu, Yule Liu, Zhen Yang, Rui Lu, Yuxiao Dong, Jie Tang, Evgeny Kharlamov
The paper introduces Regularized Latent Dynamics Prediction (RLDP), a method that adds orthogonality regularization to self‑supervised next‑state prediction in latent space. RLDP maintains feature diversity, matching or surpassing complex representation learning approaches for zero‑shot reinforcement learning. It also performs robustly in low‑coverage data settings where prior methods fail.
By Pranaya Jajoo, Harshit Sikchi, Siddhant Agarwal, Amy Zhang, Scott Niekum, Martha White
The paper introduces AgentDiff, a metric that quantifies how much LLM agents’ answers differ when inputs are altered by meaning‑bearing rewrites (paraphrases, synonym substitutions) versus presentation changes (reordering, formatting, distractors). Across 68 model–benchmark–scaffold combinations involving ten LLMs and over 1,500 questions, meaning‑bearing rewrites consistently produce a roughly 20‑percentage‑point higher inconsistency rate than presentation changes, a gap that persists across severity proxies and remains significant even outside the Qwen family. Trace analysis reveals that meaning‑bearing rewrites preserve the first action but reduce thought similarity from the second step onward, extending the divergence cascade—a phenomenon termed “stealth divergence.”
By Liyun Zhang, Jiayi Guo
The paper introduces Routed Graph Handoff, a lightweight LLM router that chooses between a typed dependency graph and natural language for each delegation in multi‑agent LLM systems. On four benchmarks with over 1,050 trajectories, the routed system matches or surpasses natural‑language‑only performance, achieving significant compression and accuracy gains. The approach requires a graph‑aware executor prompt, and an oracle analysis suggests further potential for execution‑time adaptive routing.
By Pratyay Banerjee, Ankit Chadha
The paper introduces a self‑evolving defense framework for large language models that uses a persistent, cross‑interaction rule memory to adapt to new jailbreak attacks. When an attack succeeds, the system abstracts the failure into a method‑level rule that captures the structural attack wrapper, allowing the rule to generalize across an entire attack family. This memory‑based adaptation operates without parameter updates, works with both open‑weight and black‑box models, and has been shown to reduce attack success rates while preserving benign utility across multiple jailbreak families.
By Tongyan Hu, Bryan Hooi
EduDial is a large-scale multi-turn teacher‑student dialogue corpus covering 345 core knowledge points and 34,250 dialogue sessions, designed around Bloom’s taxonomy and ten questioning strategies such as situational, ZPD, and metacognitive questioning. The dataset includes differentiated teaching strategies for students at varying cognitive levels to provide targeted guidance. Using EduDial, the authors trained EduDial‑LLM 32B and introduced an 11‑dimensional evaluation framework that measures teaching quality and content quality, showing that most mainstream LLMs struggle with student‑centered teaching while EduDial‑LLM outperforms all baselines across all metrics.
By Shouang Wei, Min Zhang, Xin Lin, Bo Jiang, Zhongxiang Dai, Kun Kuang
The paper introduces Agent Evolving Learning (AEL), a two‑timescale framework that dynamically evolves an LLM agent’s memory‑retrieval harness in open‑ended environments. A fast Thompson‑Sampling bandit selects among retrieval policies each episode, while a slower LLM reflection diagnoses performance drops and injects new policies when the current set plateaus. AEL outperforms ten self‑improving and non‑LLM baselines on a sequential portfolio benchmark, boosting Sharpe ratio by 27% and achieving significant accuracy gains on a support‑ticket routing stream.
By Wujiang Xu, Jiaojiao Han, Minghao Guo, Kai Mei, Xi Zhu, Han Zhang, Dimitris N. Metaxas
The paper evaluates additive activation steering in chat and agent contexts, showing that the commonly used gain ratio (Δ_agent/Δ_chat) fails to reliably indicate potency and efficacy across multiple models and dose-response cells. By replacing the gain with a location metric, dEC50 (difference in EC50 between agent and chat), the authors demonstrate a more robust, two‑sided measure that consistently captures cross‑context shifts. The study also reports several refuted and unanswered claims, emphasizing that a single operating point cannot distinguish between displacement and gain effects.
By Lucas Pinto
The Planetary Prediction Engine (PPE) is an autonomous AI system that transforms natural-language queries into end-to-end geospatial predictions. It automatically retrieves and fuses multimodal datasets from open-web and Earth observation sources, incorporates foundation model embeddings, and searches task‑specific model families with overfitting safeguards. Across multiple domains, PPE outperforms state‑of‑the‑art baselines, improving regression metrics for CDC health indicators, FEMA risk indices, and the Social Vulnerability Index, doubling accuracy for Nigerian food security indicators, and achieving higher recall in Ebola outbreak nowcasting.
By Evelyn Ma, Rama Kumar Pasumarthi, Kishwar Shafin, Mandar Sharma, Mimi Sun, Hamed Sadeghi, Dav M. Ebengo, Mbulayi Onesime, Rouslan Solomakhin, John Wamburu, William Ogallo, Aisha Walcott-Bryant, Sanxing Chen, Arbaaz Muslim, Yael Mayer, Ronald Ho, Roy Lee, Ruth Alcantara, Abdoulaye Diack, Monica Bharel, Lambert Rosique, Jeremy Amez-Droz, Christopher Haire, James Manyika, Yossi Matias, Niv Efron, Gautam Prasad, Shravya Shetty
LaGen is an autoregressive framework that generates long‑horizon LiDAR scenes frame by frame, using a single‑frame input and bounding‑box conditions to produce high‑fidelity 4D scenes. It introduces a scene decoupling estimation module for better object‑level interaction and a noise modulation module to reduce error accumulation over time. Evaluations on the nuScenes dataset show that LaGen outperforms existing methods, especially on later frames.
By Sizhuo Zhou, Xiaosong Jia, Fanrui Zhang, Junjie Li, Juyong Zhang, Yukang Feng, Jianwen Sun, Songbur Wong, Junqi You, Junchi Yan
The paper introduces pFedMARL, a federated learning framework that uses multi‑agent reinforcement learning with TD3 to dynamically adjust client contributions and personalize models. It applies a server‑side agent to optimize global aggregation and client‑side agents to balance global and local updates, eliminating the need for pre‑training. Experiments on a semi‑supervised audio spectrogram transformer show that pFedMARL outperforms or matches FedAvg, Ditto, and local training across various non‑IID settings and against adversarial clients, improving accuracy, robustness, and fairness.
By Rene Glitza, Luca Becker, Rainer Martin
PIVOT is a new multi‑trajectory dataset and evaluation framework that captures real‑world scenes with diverse camera paths, preserving both sensor‑derived measured poses and COLMAP‑optimized poses along with calibrated and optimized intrinsics. It defines three benchmark families—seen vs. unseen trajectory generalization, measured vs. optimized pose sensitivity, and calibrated vs. optimized intrinsics sensitivity—and introduces a directed pose‑space Chamfer distance to assess pose coverage. The first version of PIVOT includes five scenes recorded with a DJI Mini 4 Pro and offers an open processing and Nerfstudio‑based evaluation toolchain, revealing a consistent quality gap between held‑out and unseen trajectories and significant sensitivity to pose source and camera intrinsics.
By Mary Raymond
arXiv:2608. 26086v1 Announce Type: new Abstract: Large language models write correct code for isolated problems but remain far weaker at autonomous machine-learning development, where an agent must revise data pipelines, models, and validation over hours of feedback, and on most competitions still finishes below strong human competitors.
By Jiarui Yan, Weiwei Sun, Sijie Li, Wenhan Li, Yiming Yang
Scorpio is an LLM serving system that optimizes for heterogeneous Service Level Objectives (SLOs) such as Time to First Token (TTFT) and Time Per Output Token (TPOT). It uses adaptive scheduling across admission control, queue management, and batch selection, featuring a TTFT Guard that reorders requests by least-deadline-first and rejects unattainable ones, and a TPOT Guard that employs VBS-based admission control and a credit-based batching mechanism. Predictive modules support both guards, and evaluations show Scorpio can increase system goodput by up to 14.4× and improve SLO adherence by up to 46.5% under high load compared to state-of-the-art baselines.
By Yinghao Tang, Tingfeng Lan, Bo Pan, Xiuqi Huang, Hui Lu, Wei Chen