PlannerForge is a unified LLM‑agent framework that covers the entire scenario‑based testing pipeline for autonomous driving systems, from scenario generation to ADS assessment, and adds ADS enhancement and benchmarking stages. It was evaluated with ten off‑the‑shelf LLMs across all tasks and five prompt conditions, achieving best‑per‑task scores between 0.88 and 1.00 and matching commercial APIs with open‑source models such as Qwen3.6:35B. The end‑to‑end chaining retains 83% of seed queries for commercial backends and 78% for open‑source, outperforming existing tools like Scenario Factory 2.0 and BM25 in natural‑language generation, attribute realization, and physically valid edits.
whyItMatters":"PlannerForge demonstrates that a single LLM‑based system can streamline and improve the fragmented scenario‑based testing workflow for autonomous driving, achieving high performance without domain‑specific fine‑tuning."
By Yuan Gao, Sebastian M\"uller, Mattia Piccinini, Marc Kaufeld, Yuchen Zhang, Finn Rasmus Sch\"afer, Qunying Song, Johannes Betz
Agentick is a unified benchmark for sequential decision‑making agents that evaluates RL, LLM, VLM, hybrid, and human agents on 37 procedurally generated tasks across six capability categories, four difficulty levels, and five observation modalities via a single Gymnasium‑compatible interface. It includes a Coding API, oracle reference policies, pre‑built SFT datasets, a composable agent harness, and a live leaderboard. An evaluation of 27 configurations and over 90,000 episodes shows no single approach dominates, with GPT‑5 mini leading overall, PPO excelling in planning and multi‑agent tasks, and the reasoning harness boosting LLM performance by 3‑10×, while ASCII observations outperform natural language.
By Roger Creus Castanyer, Pablo Samuel Castro, Glen Berseth
GRASP is a multi-stage planning framework that improves the reliability of large language models on complex tasks. It separates planning into three specialized modules—GenPlan for global macro-guidelines, RevPlan for exploring localized strategies, and VerPlan for multi-criteria evaluation—allowing context isolation and strict macro-regularization. Experiments show GRASP outperforms direct LLM planners by significant margins on datasets such as Natural Plan Calendar Scheduling, ZebraLogic, and SciBench Math, and it mitigates performance collapse in multi-task and dual-task settings.
By Arunabh Srivastava (Amir), Mohammad A. (Amir), Khojastepour, Srimat Chakradhar, Sennur Ulukus
arXiv:2608. 05519v1 Announce Type: new Abstract: Agent benchmarks usually measure task completion and treat resource use as an auxiliary statistic.
By Jie Wu, Ming Gong, Feixiang Cheng, Qinqin Zhao
arXiv:2607. 29577v1 Announce Type: new Abstract: Games and simulators make valuable benchmarks by turning decisions into measurable outcomes, but many current suites under-test rules-rich tactical reasoning: the ability to choose well when geometry, timing, resources, objectives, and rule interactions all matter at once.
By Ismayil Ismayilov, Atakan Kara, Kaan Oktay
The paper introduces RuVerBench, a benchmark with 2,458 instances for evaluating the reliability of Large Language Models acting as judges (LaaJ) in verifying rubric compliance within agentic scenarios such as deep research and agentic coding. It systematically meta‑evaluates frontier LLMs, revealing that even the most advanced models perform well yet still produce substantial noise. The study also examines how prompt design, batching, and majority voting affect verification accuracy, noting that weaker models are more prompt‑sensitive, batched verification trades accuracy for efficiency, and majority voting offers diminishing returns.
By Yangda Peng, Yunjia Qi, Haotian Xia, Guanzhong He, Xintong Shi, Richeng Xuan, Songyuanyi Lu, Yixian Liu, Zhichao Hu, Yuhong Liu, Hao Peng
GRASP is a multi-stage planning framework that separates planning into specialized modules: GenPlan for global macro-guidelines, RevPlan for exploring localized strategies, and VerPlan for multi-criteria evaluation. This strategy-aware approach yields state‑of‑the‑art accuracy on diverse datasets, outperforming direct LLM planners by up to 30.8% on ZebraLogic and reducing multi‑task degradation. GRASP’s context isolation and macro‑regularization also give it a 14.5% edge over frontier reasoning models like GPT‑5‑mini.
arXiv:2608.23035v1 Announce Type: new
Abstract: As on-device LLM agents evolve into personal copilots, the mobile operating system has become a key testbed for this paradigm, making rigorous capabili...
By Yi Zhu, Xiongwei Wu, Qiyi Wang, Tingyu Qu, Jiajun Liu, Sihan Cao, Long Chen, Weigao Sun, Feida Zhu, Yiran Zhong, Steven Hoi
AgentAudit is an open, extensible framework that evaluates the full lifecycle of AI agents, assessing planning, tool selection, execution, memory, and reasoning across ten dimensions such as instruction integrity, security, and alignment. Unlike existing benchmarks that focus on single aspects, AgentAudit analyzes the entire execution trace to attribute failures to specific stages. The framework was tested on five large language models, revealing significant differences in trustworthiness even among models with similar task‑completion performance.
By Shrey Nag, Sachita, Abhishek Kumar Singh, Lipi Goel, Rajeshwar Singh Janwar
MoMHa is a system that optimizes large language model harnesses across three objectives—accuracy, behavioural safety, and token cost—using a single‑phase joint‑reward proposer. It outperforms alternative strategies on seventeen domains, including synthetic suites and real‑world benchmarks, achieving higher joint scores and better safety while reducing token usage. The approach demonstrates that multi‑objective harness design can transfer effectively to unseen models and tasks.
By Subhojyoti Mukherjee, Md Mehrab Tanjim
arXiv:2606. 09118v1 Announce Type: new Abstract: As LLM capabilities advance rapidly, the evaluation methods used to assess them increasingly lag behind.
By Sushant Mehta, Liudas Panavas, Edwin Chen
arXiv:2602. 05459v2 Announce Type: replace Abstract: Offline goal-conditioned reinforcement learning (GCRL) is typically benchmarked by the best tuned success rate of each method.
By Jan Malte T\"opperwien, Aditya Mohan, Marius Lindauer