The paper presents a formal framework for applying spiking neural networks to satellite onboard computing, deriving closed‑form energy models based on network sparsity, firing rate, and fan‑out. It introduces three novel spike‑based algorithms—SKCF for attitude estimation, TCAD for health monitoring, and SAMC for adaptive communication—alongside analytical complexity bounds. A quantitative Energy‑Accuracy‑Latency surface, populated with silicon data from Loihi and SpiNNaker, demonstrates that neuromorphic solutions can reduce energy consumption by 8 to 47 times compared to GPU baselines while keeping inference accuracy within 2.1% for typical satellite workloads.
By Suresh Kumar TP, Vikas Agnihotri
The paper introduces GraphTok, a GPU-based tokenizer that executes the entire tokenization pipeline as a single CUDA graph, matching reference token IDs for models such as GPT‑2, Llama 3, Qwen 2.5, DeepSeek‑V3, and Gemma 3. GraphTok achieves significant speedups, encoding one million GPT‑2 tokens in 1.7–3.2 ms and outperforming the leading CPU tokenizer gigatoken by 4.7–6.0× on Llama 3. In a vLLM deployment with cached prefixes, it reduces mean time‑to‑first‑token by 2.1–2.4× and dramatically increases request throughput, sustaining 518 req/s of 121k‑token traffic versus 296 with gigatoken.
By Mohammad Siavashi, Pascal Schwinte, Ali Banaei, Gerald Q. Maguire Jr., Marco Chiesa, Dejan Kosti\'c
The paper presents a fully data-driven method for optimal control of nonlinear control‑affine systems modeled as stochastic diffusions. It embeds state probability densities into a reproducing kernel Hilbert space to learn Markov transition operators, which are then used in convex operator‑theoretic Hamilton‑Jacobi‑Bellman recursions that scale linearly with state dimensionality. Numerical experiments show the approach can solve diverse nonlinear control tasks, such as depth regulation of an autonomous underwater vehicle.
By Nicolas Hoischen, Petar Bevanda, Stefan Sosnowski, Sandra Hirche, Boris Houska
SWE‑Duel proposes a self‑scaling benchmark for coding agents, where tasks and oracles are generated at evaluation time to prevent data leakage and adapt difficulty to participant performance. In the arena, two agents take turns as Red and Blue: Red proposes a feature, injects a subtle bug, and must repair it blind, while Blue reviews the pull request without knowing the bug and must pass all tests. A tournament of 11 model‑harness participants on 15 repositories across five languages showed GPT‑5.5/Codex CLI as the top solver and Claude Opus 4.8/mini‑swe‑agent as the top task author, with newer model versions producing harder tasks for older ones.
By Wei Minn, Zhensu Sun, Lwin Khin Shar, David Lo
GenCoord is a decentralized protocol for embodied language‑model agents that turns the division of work into the communication message itself. By pairing a sender’s skill path with a partner’s continuation in a short domain‑specific language, agents create commitments that link delivery to subsequent work, allowing fulfillment to be verified separately from task success. Experiments in Minecraft show that GenCoord matches free‑form communication performance while cutting median commitment time by 68.2 % and reducing output tokens, demonstrating that structured, work‑based messages can improve efficiency in embodied cooperation.
By Junning Zhu, Peng He, Haohan Yuan, Jianpeng Liang
The study examines whether the built‑in FAIL_TO_PASS (F2P) tests in repository‑level issue‑resolution benchmarks can be used to evaluate the runtime performance of patches that pass correctness tests. Using four advanced language models paired with SWE‑agent, the authors measured instruction counts and wall‑clock times for 262 cases and applied Welch ANOVA to detect performance differences. They found that many patches exhibit significant runtime disparities, that performance rankings often contradict issue‑resolution rates, and that some architectural regressions are revealed only through performance analysis.
By Han Peng, Xing Hu, Feifei Niu, Huan Li, Xin Xia
The study evaluates multi-harness reinforcement learning (RL) for coding agents by training two Qwen3-8B RL variants—one with rewards normalized within each harness and one across harnesses—using identical replay data. Results show modest gains over the supervised fine-tuning baseline (+0.48 and +0.77 percentage points for within and cross normalization, respectively), but no consistent advantage of cross normalization on unseen harnesses. Performance differences across harnesses are much larger than the training gains, indicating that harness dependence remains a significant challenge.
By Chenqian Le, Jiayi Cheng, Qijia He, Runhao Li, Yinghao Li, Xupeng Chen
The paper investigates how to encourage agents in a multi‑agent game to report unsafe tasks when rewards are shared. It shows that a fully shared policy can be drawn toward silence even when reporting is optimal, and derives audit conditions that guarantee policy‑gradient updates lead to universal reporting. Experiments with PPO on 24 witness graphs demonstrate that grouping witnesses into separate networks reduces unsafe task completion by over 30 percentage points compared to shuffled groups, while a fully shared network only succeeds under a common action draw.
By Avyay M. Casheekar, Hariganesh Tangirala
FPD-Drive is an end‑to‑end autonomous driving planner that uses future‑privileged multi‑target distillation to transfer diverse, high‑quality trajectory targets from a teacher model with access to future observations to a student policy. It also introduces safety‑aware score correction to improve candidate selection by adjusting rankings between safe and unsafe trajectories. The approach achieves state‑of‑the‑art performance on NAVSIM v1 and v2 benchmarks and demonstrates strong zero‑shot closed‑loop results on HUGSIM.
By Li-Heng Chen, Qifeng Chen, Chengwei Wu, Yingjie Zhang, Sheng Yang, Hongbo Fu, Shaoqing Ren
SOAR (Source-Aware Reinforcement learning) is a new paradigm that transforms monolithic reasoning into source-explicit trajectories for search agents. It disentangles external evidence ($E_{ext}$) and internal knowledge ($E_{int}$) into parallel, observable streams via a Source-Aware Workflow and uses a Source-Consistent Reward to guide policy evolution. Evaluations show that SOAR improves performance on 3B and 7B backbones by 7.3% and 5.4% respectively, while reducing evidence fabrication by 74.8%.
By Shengze Huang, Zihao Tang, Jinluan Yang, Xin Yu, Zhenghao Lin, Xiao Liu, Zengxuan Wen, Yeyun Gong, Haizhen Huang, Weiwei Deng, Feng Sun, Qi Zhang, Fei Wu, Kun Kuang
The paper introduces POrchestra, a proactive agentic orchestration framework that lets sub‑agents request harness adjustments during execution, enhancing adaptability to dynamic environments. It also proposes On‑Policy Orchestration Distillation (OPOD), which uses hindsight feedback from sub‑agents’ requests to provide targeted improvement signals for training the orchestration policy. Experiments on three long‑horizon benchmarks show that POrchestra boosts task success by about 9.1% over the best baseline, and with OPOD the improvement rises to 10.0%, while also improving cost‑performance trade‑offs.
By Xueyang Feng, Chaozhuo Li, Xiaohe Bo, Jiakai Tang, Chenxi Li, Qiwei Ye, Xu Chen
State-OPD introduces a state‑grounded on‑policy distillation framework that separates the objective task and action‑observation record from the student’s internal reasoning trajectory. By removing prior student reasoning and resampling complete reasoning‑action responses at high‑entropy states, the method improves teacher supervision and action diversity. Evaluations on ALFWorld, ScienceWorld, and WebShop show State‑OPD achieves the highest macro‑average success rates for Qwen3‑1.7B and Qwen3‑4B students, even surpassing a larger teacher model on certain tasks.
By Zhentao Zhang, Yanzhao Zheng, Yanlin Liu, Yuanqiang Yu, Tianze Xu, Chao Ma, Jihuai Zhu, Lelun Wu, Baohua Dong, Hangcheng Zhu, Ruohui Huang, Gang Yu
AgentTune is a source‑available Python library that offers a unified training lifecycle for tool‑using large language model agents. It introduces a two‑tier EventLog representation to preserve agent trajectories and enable consistent training, evaluation, and auditing. The framework supplies a shared runtime, reinforcement learning for tool‑calling, agentic distillation, retrieval‑augmented generation, decision workflows, evaluation utilities, and a path toward self‑healing agents, all exposed through composable interfaces for flexible combination of strategies and methods.
By R E Zera Marveen Lyngkhoi, Abhivansh Gupta, Vidushee Vats, Ram Mohan Rao Kadiyala, Vinay Kumar Sankarapu, Pratinav Seth
LongHarness is a model‑agnostic harness that explicitly manages the interaction between acquired evidence and evolving conclusions for long‑context reasoning over corpora larger than a model’s context window. Its Evidence‑Conclusion Coevolution mechanism uses unresolved conclusions to steer evidence retrieval, while new evidence refines those conclusions, and its Evidence Lifecycle Management keeps evidence current and traceable. Evaluations on OfficeQA Pro, AA‑LCR, and LongBench v2 show substantial performance gains over direct reasoning and state‑of‑the‑art agent harnesses, with GLM‑5.3‑Flash exact match improving from 15.8% to 60.2% on V1 and from 4.4% to 61.1% on V2.
By Songtao Huang, Zhijie Zhong, Xiaohan He, Shuaiyu Zhang, Zhiyin Yu, Li Kang, Heng Zhou, Yutao Fan, Xiangchao Yan, Bo Zhang, Shiyang Feng
The paper proposes a holistic architectural and algorithmic framework for sustainable cloud‑fog‑edge systems in smart buildings, emphasizing real‑time responsiveness and reduced environmental impact. It validates resilient data acquisition using affordable Arduino hardware and secure persistence solutions, introduces the lightweight KOptim orchestration framework for dynamic SLA management, and adapts the Generalized Hebbian Algorithm for PCA on low‑power microcontrollers with an iterative learning strategy that preserves data diversity. Experimental results on high‑dimensional sensor data demonstrate that this approach converges to quality comparable to full‑memory processing while operating within strict memory constraints.
By Mamadou Sow, Christophe Cerin
The paper introduces an operator-theoretic approach to offline reinforcement learning for continuous-time stochastic systems. By linking reinforcement learning to the Hamilton‑Jacobi‑Bellman equation, it proposes a dynamic programming recursion that represents the world model through the infinitesimal generator of controlled diffusion processes learned in a reproducing kernel Hilbert space. The authors establish global convergence of the value function under small generator‑approximation error and provide explicit finite‑sample guarantees tied to system properties such as smoothness and stability.
By Nicolas Hoischen, Petar Bevanda, Max Beier, Stefan Sosnowski, Boris Houska, Sandra Hirche
The paper investigates whether restoring full sandbox snapshots can aid coding agents in recovering from failures. By implementing a Go‑Explore‑style search over complete Daytona snapshots, the authors find that higher and more efficient solve rates arise when progress is stored outside the repository—such as in databases, services, and other runtime state. The study suggests that full‑snapshot search is most beneficial for long‑horizon coding tasks that span multiple services whose state cannot be reconstructed from a simple diff.
By Daniel Thi Graviet, Tanmay Shah
StereoEngine introduces two methods for generating realistic synthetic stereo data for robotic manipulation. The first method warps a monocular video into a right view using depth estimation and a video diffusion model to inpaint disoccluded regions, while the second method renders stereo pairs directly in simulation. These approaches outperform existing baselines in stereo quality and enable improved bimanual manipulation policies when used for mid‑training or task‑specific fine‑tuning.
By Jason Chen, I-Chun Arthur Liu, Gaurav Sukhatme, Daniel Seita
The paper investigates how snow affects LiDAR-based object detection, finding that detection accuracy actually improves with higher snow intensity. This counter‑intuitive result is explained by an increase in object point count due to snow accumulation, which outweighs the noise introduced by falling snow. The authors argue that broad weather severity bins can mislead robustness assessments unless atmospheric noise and surface accumulation effects are disentangled.
By Yiming Yang, Jeremy P. Bos
The paper introduces SOAP, a method for attributing failures in large language model agents without requiring step-level labels or a separate set of successful trajectories. SOAP builds a spectral reference space from LLM representations to assign base error scores to each step, then uses attention-based soft dependencies to propagate downstream error signals back to their likely upstream causes. Experiments demonstrate that SOAP outperforms existing baselines in accurately identifying decisive steps responsible for task failures.
By Do Nguyen-Thanh, Junlin Fang, Bing Li, Zhen Fang, Sean Du