Daily Papers

People are increasingly delegating tasks to AI agents, and those agents are increasingly encountering other people's agents over shared resources such as a codebase, a calendar, or a budget. When each agent acts for a different user with different goals, coordination often fails, and the group ends up worse off than if a single agent had acted for everyone. We study this multi-user, multi-agent setting across five frontier models and 77 scenarios in four environments: an API key environment in which agents share a compute budget, a clinic in which they share a calendar, a personal assistant environment in which they share a group order or booking, and a merge queue in which they share a release cutoff. In each scenario, we compare a single agent that serves every user (a coordinator) to a team in which each agent serves one user, with and without a communication channel between the agents. Teams deliver worse group outcomes than the coordinator in every environment: without a channel, they completely collapse in two environments, and even with one, coordination overhead creates substantial gaps. For example, in the personal assistant environment, the coordinator fulfills a targeted user request about twice as often as teams. We identify distinct behaviors associated with this poor group-level performance, including stalling as teams grow, overriding each other's actions, and fabricating claims. We find effective but environment-specific mitigations, such as a team lead, explicit procedural instructions, and a platform check that makes an agent read its peers' messages before committing. We will release the API key, clinic, and personal assistant environments as MAMUBench, comprising 74 scenarios for evaluating multi-user, multi-agent coordination.

We introduce Berkeley Humanoid, a reliable and low-cost mid-scale humanoid research platform for learning-based control. Our lightweight, in-house-built robot is designed specifically for learning algorithms with low simulation complexity, anthropomorphic motion, and high reliability against falls. The robot's narrow sim-to-real gap enables agile and robust locomotion across various terrains in outdoor environments, achieved with a simple reinforcement learning controller using light domain randomization. Furthermore, we demonstrate the robot traversing for hundreds of meters, walking on a steep unpaved trail, and hopping with single and double legs as a testimony to its high performance in dynamical walking. Capable of omnidirectional locomotion and withstanding large perturbations with a compact setup, our system aims for scalable, sim-to-real deployment of learning-based humanoid systems. Please check http://berkeley-humanoid.com for more details.

The rapid evolution of autonomous, agentic artificial intelligence within financial services has introduced an existential architectural crisis: large language models (LLMs) are probabilistic, non-deterministic systems operating in domains that demand absolute, mathematically verifiable compliance guarantees. Existing guardrail solutions -- including NVIDIA NeMo Guardrails and Guardrails AI -- rely on probabilistic classifiers and syntactic validators that are fundamentally inadequate for enforcing complex multi-variable regulatory constraints mandated by the SEC, FINRA, and OCC. This paper presents the Lean-Agent Protocol, a formal-verification-based AI guardrail platform that leverages the Aristotle neural-symbolic model developed by Harmonic AI to auto-formalize institutional policies into Lean 4 code. Every proposed agentic action is treated as a mathematical conjecture: execution is permitted if and only if the Lean 4 kernel proves that the action satisfies pre-compiled regulatory axioms. This architecture provides cryptographic-level compliance certainty at microsecond latency, directly satisfying SEC Rule 15c3-5, OCC Bulletin 2011-12, FINRA Rule 3110, and CFPB explainability mandates. A three-phase implementation roadmap from shadow verification through enterprise-scale deployment is provided.

In recent years, the integration of Large Language Models (LLMs) into recommender systems has garnered interest among both practitioners and researchers. Despite this interest, the field is still emerging, and the lack of open-source R&D platforms may impede the exploration of LLM-based recommendations. This paper introduces OpenP5, an open-source platform designed as a resource to facilitate the development, training, and evaluation of LLM-based generative recommender systems for research purposes. The platform is implemented using encoder-decoder LLMs (e.g., T5) and decoder-only LLMs (e.g., Llama-2) across 10 widely recognized public datasets, catering to two fundamental recommendation tasks: sequential and straightforward recommendations. Recognizing the crucial role of item IDs in LLM-based recommendations, we have also incorporated three item indexing methods within the OpenP5 platform: random indexing, sequential indexing and collaborative indexing. Built on the Transformers library, the platform facilitates easy customization of LLM-based recommendations for users. OpenP5 boasts a range of features including extensible data processing, task-centric optimization, comprehensive datasets and checkpoints, efficient acceleration, and standardized evaluations, making it a valuable tool for the implementation and evaluation of LLM-based recommender systems. The open-source code and pre-trained checkpoints for the OpenP5 library are publicly available at https://github.com/agiresearch/OpenP5.

In this paper, we present JADE, a targeted linguistic fuzzing platform which strengthens the linguistic complexity of seed questions to simultaneously and consistently break a wide range of widely-used LLMs categorized in three groups: eight open-sourced Chinese, six commercial Chinese and four commercial English LLMs. JADE generates three safety benchmarks for the three groups of LLMs, which contain unsafe questions that are highly threatening: the questions simultaneously trigger harmful generation of multiple LLMs, with an average unsafe generation ratio of 70% (please see the table below), while are still natural questions, fluent and preserving the core unsafe semantics. We release the benchmark demos generated for commercial English LLMs and open-sourced English LLMs in the following link: https://github.com/whitzard-ai/jade-db. For readers who are interested in evaluating on more questions generated by JADE, please contact us. JADE is based on Noam Chomsky's seminal theory of transformational-generative grammar. Given a seed question with unsafe intention, JADE invokes a sequence of generative and transformational rules to increment the complexity of the syntactic structure of the original question, until the safety guardrail is broken. Our key insight is: Due to the complexity of human language, most of the current best LLMs can hardly recognize the invariant evil from the infinite number of different syntactic structures which form an unbound example space that can never be fully covered. Technically, the generative/transformative rules are constructed by native speakers of the languages, and, once developed, can be used to automatically grow and transform the parse tree of a given question, until the guardrail is broken. For more evaluation results and demo, please check our website: https://whitzard-ai.github.io/jade.html.

Automatically verifying climate-related claims against scientific literature is a challenging task, complicated by the specialised nature of scholarly evidence and the diversity of rhetorical strategies underlying climate disinformation. ClimateCheck 2026 is the second iteration of a shared task addressing this challenge, expanding on the 2025 edition with tripled training data and a new disinformation narrative classification task. Running from January to February 2026 on the CodaBench platform, the competition attracted 20 registered participants and 8 leaderboard submissions, with systems combining dense retrieval pipelines, cross-encoder ensembles, and large language models with structured hierarchical reasoning. In addition to standard evaluation metrics (Recall@K and Binary Preference), we adapt an automated framework to assess retrieval quality under incomplete annotations, exposing systematic biases in how conventional metrics rank systems. A cross-task analysis further reveals that not all climate disinformation is equally verifiable, potentially implicating how future fact-checking systems should be designed.

Large language models are redefining software engineering by implementing AI-powered techniques throughout the whole software development process, including requirement gathering, software architecture, code generation, testing, and deployment. However, it is still difficult to develop a cohesive platform that consistently produces the best outcomes across all stages. The objective of this study is to develop a unified platform that utilizes multiple artificial intelligence agents to automate the process of transforming user requirements into well-organized deliverables. These deliverables include user stories, prioritization, and UML sequence diagrams, along with the modular approach to APIs, unit tests, and end-to-end tests. Additionally, the platform will organize tasks, perform security and compliance, and suggest design patterns and improvements for non-functional requirements. We allow users to control and manage each phase according to their preferences. In addition, the platform provides security and compliance checks following European standards and proposes design optimizations. We use multiple models, such as GPT-3.5, GPT-4, and Llama3 to enable to generation of modular code as per user choice. The research also highlights the limitations and future research discussions to overall improve the software development life cycle. The source code for our uniform platform is hosted on GitHub, enabling additional experimentation and supporting both research and practical uses. \end

Data policies for reinforcement learning with verifiable rewards (RLVR) determine which rollouts are used, how strongly they are weighted, and which domains contribute to subsequent training batches. We introduce DataFlex-RL, an evaluation platform for comparing these choices under a common GRPO recipe. Our primary experiment evaluates 13 configurations across 12 matched seeds using Qwen2.5-7B-Base and 12 mathematics, logic, and science benchmarks. Uniform GRPO improves the domain-balanced average accuracy by 7.76 percentage points over the untrained checkpoint. None of the eight rollout-selection or reweighting methods achieves a paired 95% confidence interval that excludes zero relative to uniform sampling, and none of the three adaptive mixtures outperforms a fixed equal mixture at the same level of precision. A corrected 12-seed extension on Llama-3.1-8B-Base places the additional methods on the same score scale as the original controls, but does not reveal a consistent winner in terms of observed mean performance. We also quantify evaluation sensitivity by rescoring nine Qwen2.5-7B-Instruct runs using a math-heavy six-benchmark summary, consisting of five mathematics benchmarks and GPQA-Diamond but no logic benchmark, and comparing it with the domain-balanced 12-benchmark summary. The resulting rankings are negatively correlated, with a correlation coefficient of -0.33, whereas summaries that retain all 12 benchmarks largely agree. Across the controlled settings studied here, changing the data policy measurably changes the training process but does not produce a reproducible improvement over uniform training.

The development of embodied agents for complex commercial environments is hindered by a critical gap in existing robotics datasets and benchmarks, which primarily focus on household or tabletop settings with short-horizon tasks. To address this limitation, we introduce MarketGen, a scalable simulation platform with automatic scene generation for complex supermarket environments. MarketGen features a novel agent-based Procedural Content Generation (PCG) framework. It uniquely supports multi-modal inputs (text and reference images) and integrates real-world design principles to automatically generate complete, structured, and realistic supermarkets. We also provide an extensive and diverse 3D asset library with a total of 1100+ supermarket goods and parameterized facilities assets. Building on this generative foundation, we propose a novel benchmark for assessing supermarket agents, featuring two daily tasks in a supermarket: (1) Checkout Unloading: long-horizon tabletop tasks for cashier agents, and (2) In-Aisle Item Collection: complex mobile manipulation tasks for salesperson agents. We validate our platform and benchmark through extensive experiments, including the deployment of a modular agent system and successful sim-to-real transfer. MarketGen provides a comprehensive framework to accelerate research in embodied AI for complex commercial applications.

SVD-based low-rank compression has become a fast-growing direction for reducing the memory and computational cost of large language models (LLMs). However, meaningful comparison across existing studies remains difficult as prior evaluations use varied benchmarks, inconsistent ratios, and diverse setups, often failing to isolate low-rank effects from auxiliary techniques. As a result, it remains unclear whether reported gains reflect method-level improvements or differences in evaluation protocol. This lack of comparability highlights the need for a unified, reproducible evaluation platform. To address this problem, we present LowRankArena, a standardized evaluation platform for SVD-based LLM compression. LowRankArena unifies task versions, uniform-precision compression budgets, comparison regimes, and inference measurements, and provides a reproducible pipeline with over 3 TiB released compressed checkpoints. Using LowRankArena, our aligned audit of five representative SVD methods reveals that prior findings are highly conditional under standardized protocols: clear leaders and performance tiers shift across backbones and keep ratios, multiple-choice accuracy can hide large perplexity degradation, and nominal low-rank savings yield workload-dependent and often limited end-to-end speedups. Our code is available at: https://github.com/Zishan-Shao/lowrankarena.git.

Proof assistants such as Lean 4 promise the paradigm of formally verified mathematics, but large-scale formalization projects have faced major barriers to entry, including the need for expertise in formal verification (as well as the underlying mathematics) and the significant time required for writing formal proofs. AI coding agents have dramatically reduced these barriers; human users can now use natural language to prompt agents to write complex proofs in Lean. This opens up the intriguing possibility of internet-scale mathematical collaboration involving both humans and AI agents, where correctness is machine-checked. To realize this possibility, we introduce Prove2Me (https://prove2.me), an open collaborative platform for formalizing mathematics. Users launch formalization "missions", to which AI agents contribute formal proofs toward completion. We designed mechanisms and a specialized harness in Prove2Me that enable large-scale collaboration so that agents can build on one another's work and freely reuse existing results. In doing so, Prove2Me aims to turn math formalization into a scalable, crowd-sourced effort open to anyone with an agent.

Two commonly-employed strategies to combat the rise of misinformation on social media are (i) fact-checking by professional organisations and (ii) community moderation by platform users. Policy changes by Twitter/X and, more recently, Meta, signal a shift away from partnerships with fact-checking organisations and towards an increased reliance on crowdsourced community notes. However, the extent and nature of dependencies between fact-checking and helpful community notes remain unclear. To address these questions, we use language models to annotate a large corpus of Twitter/X community notes with attributes such as topic, cited sources, and whether they refute claims tied to broader misinformation narratives. Our analysis reveals that community notes cite fact-checking sources up to five times more than previously reported. Fact-checking is especially crucial for notes on posts linked to broader narratives, which are twice as likely to reference fact-checking sources compared to other sources. In conclusion, our results show that successful community moderation heavily relies on professional fact-checking.

Recent large language models such as Gemini-1.5, DeepSeek-V3, and Llama-4 increasingly adopt Mixture-of-Experts (MoE) architectures, which offer strong efficiency-performance trade-offs by activating only a fraction of the model per token. Yet academic researchers still lack a fully open, end-to-end MoE platform for investigating scaling, routing, and expert behavior. We release FLAME-MoE, a completely open-source research suite composed of seven decoder-only models, ranging from 38M to 1.7B active parameters, whose architecture--64 experts with top-8 gating and 2 shared experts--closely reflects modern production LLMs. All training data pipelines, scripts, logs, and checkpoints are publicly available to enable reproducible experimentation. Across six evaluation tasks, FLAME-MoE improves average accuracy by up to 3.4 points over dense baselines trained with identical FLOPs. Leveraging full training trace transparency, we present initial analyses showing that (i) experts increasingly specialize on distinct token subsets, (ii) co-activation matrices remain sparse, reflecting diverse expert usage, and (iii) routing behavior stabilizes early in training. All code, training logs, and model checkpoints are available at https://github.com/cmu-flame/FLAME-MoE.

We report on the first large-scale mixture-of-experts (MoE) pretraining study on pure AMD hardware, utilizing both MI300X GPUs with Pollara interconnect. We distill practical guidance for both systems and model design. On the systems side, we deliver a comprehensive cluster and networking characterization: microbenchmarks for all core collectives (all-reduce, reduce-scatter, all-gather, broadcast) across message sizes and GPU counts on Pollara. To our knowledge, this is the first at this scale. We further provide MI300X microbenchmarks on kernel sizing and memory bandwidth to inform model design. On the modeling side, we introduce and apply MI300X-aware transformer sizing rules for attention and MLP blocks and justify MoE widths that jointly optimize training throughput and inference latency. We describe our training stack in depth, including often-ignored utilities such as fault-tolerance and checkpoint-reshaping, as well as detailed information on our training recipe. We also provide a preview of our model architecture and base model - ZAYA1 (760M active, 8.3B total parameters MoE) - which will be further improved upon in forthcoming papers. ZAYA1-base achieves performance comparable to leading base models such as Qwen3-4B and Gemma3-12B at its scale and larger, and outperforms models including Llama-3-8B and OLMoE across reasoning, mathematics, and coding benchmarks. Together, these results demonstrate that the AMD hardware, network, and software stack are mature and optimized enough for competitive large-scale pretraining.

Room acoustics analysis plays a central role in architectural design, audio engineering, speech intelligibility assessment, and hearing research. Despite the availability of standardized metrics such as reverberation time, clarity, and speech transmission index, accessible tools that combine rigorous signal processing with intuitive visualization remain scarce. This paper presents AcoustiVision Pro, an open-source web-based platform for comprehensive room impulse response (RIR) analysis. The system computes twelve distinct acoustic parameters from uploaded or dataset-sourced RIRs, provides interactive 3D visualizations of early reflections, generates frequency-dependent decay characteristics through waterfall plots, and checks compliance against international standards including ANSI S12.60 and ISO 3382. We introduce the accompanying RIRMega and RIRMega Speech datasets hosted on Hugging Face, containing thousands of simulated room impulse responses with full metadata. The platform supports real-time auralization through FFT-based convolution, exports detailed PDF reports suitable for engineering documentation, and provides CSV data export for further analysis. We describe the mathematical foundations underlying each acoustic metric, detail the system architecture, and present preliminary case studies demonstrating the platform's utility across diverse application domains including classroom acoustics, healthcare facility design, and recording studio evaluation.

Computer-using agents (CUAs) are advancing rapidly across the digital world. A CUA trajectory records the agent's actions, states, and reasoning. Verifying whether it fulfilled the task instruction is central to CUA evaluation, data curation, and reinforcement learning. Neither human-written verifiers nor human annotators can provide such verification at scale, so the field increasingly turns to vision-language models (VLMs) as judges of CUA trajectories. But a fundamental question has long gone unexamined: are these VLM judges reliable enough? To study it systematically, we introduce OSReward, a realistic, high-quality benchmark that evaluates VLM judges on CUA trajectories. The trajectories come from diverse agent backbones executing human-verified instructions across platforms, then rigorously labeled with ground-truth verdicts through multi-stage human annotation. Building on it, we derive OSReward-Hard, a challenge set concentrating genuinely hard cases, and OSReward-Multi for fine-grained efficiency and alignment scoring. The most comprehensive evaluation of VLM judges to date finds even state-of-the-art models fall short of an ideal judge, sharing a systematic leniency bias that mislabels failed runs as successes. The few reliable enough to trust are too expensive to run at scale, while affordable open models trail far behind. To close this gap, we construct and release OS-Shepherd-100K, an open corpus of reasoning-annotated trajectory judgments for the CUA community. On it, we train OS-Shepherd (9B and 35B), open reward models that supply low-cost, stable, and reliable reward signals, matching commercial judges at 30-60% lower cost than the frontier. Extensive analyses further inform the design of reliable CUA reward at scale. Our code, benchmark, dataset, and model checkpoints are available at https://os-copilot.github.io/OSReward-Home/.

Online reinforcement learning (RL) with verifiable rewards has driven rapid progress in mathematics, coding, and tool use, but scaling this success to GUI agents remains challenging due to the high cost of interactive environments and the scarcity of realistic training tasks. We introduce MAI-Gym, an open all-in-one platform that unifies GUI agent evaluation, data collection, and online RL training across mobile, computer, and web environments. Through a shared rollout pipeline, MAI-Gym ensures consistent agent behavior across evaluation and training, supports ten established benchmarks, and scales to 20,000 parallel environment instances. Beyond scalable infrastructure, MAI-Gym enables automated synthesis of diverse GUI applications and challenging tasks, including resettable mocked apps and coherent cross-app worlds with execution-validated task verifiers. Using MAI-Gym, we train MAI-Gym-9B, a unified 9B agent for mobile and computer use, achieving state-of-the-art performance among open GUI agents under 10B parameters: 78.4% on MobileWorld and 76.2% on OSWorld-Verified. Importantly, RL in a mocked train-ticketing app transfers to its real-world counterpart, improving task success on a phys- ical phone from 28.8% to 48.1%. We release the framework, synthesis skills, mocked apps, 2,000 verified tasks, and the model checkpoint to support repro- ducible research and further advances in scalable GUI agent training.

High-quality structured organic reaction data are essential for developing artificial intelligence for chemistry (AI4Chem), yet much of this knowledge remains dispersed across patent text, images, and reaction schemes. We present DianShi-RxnDB, a large-scale, fine-grained organic reaction data platform built via a fully automated extraction and normalization pipeline integrating patent text, images, and reaction schemes. Its corpus covers organic synthesis patents from the USPTO and EPO published between 1976 and 2025, yielding approximately 24 million reaction instances, of which approximately 14.8 million (61.7%) pass automated qualification checks. Each instance represents a specific single-step experiment recording participants, roles, quantities, temperatures, reaction times, yields, experimental procedures, and provenance links to source patents. In a manual evaluation of 1,300 sampled qualified instances, the micro-averaged field-level accuracy was 92.95%. A matched comparison with Pistachio further indicated advantages in deduplicated record counts, representation granularity, and field-level exact agreement. The platform provides a Web research workbench for searching, filtering, comparing, and source-verifying records, and a Model Context Protocol (MCP) service offering AI agents composable structured retrieval tools. DianShi-RxnDB is available at https://dianshi.opendatalab.org.cn/ .

The standard methodology of evaluating large language models (LLMs) based on static pairs of inputs and outputs is insufficient for developing assistants: this kind of assessments fails to take into account the essential interactive element in their deployment, and therefore limits how we understand language model capabilities. We introduce CheckMate, an adaptable prototype platform for humans to interact with and evaluate LLMs. We conduct a study with CheckMate to evaluate three language models~(InstructGPT, ChatGPT, and GPT-4) as assistants in proving undergraduate-level mathematics, with a mixed cohort of participants from undergraduate students to professors of mathematics. We release the resulting interaction and rating dataset, MathConverse. By analysing MathConverse, we derive a preliminary taxonomy of human behaviours and uncover that despite a generally positive correlation, there are notable instances of divergence between correctness and perceived helpfulness in LLM generations, amongst other findings. Further, we identify useful scenarios and existing issues of GPT-4 in mathematical reasoning through a series of case studies contributed by expert mathematicians. We conclude with actionable takeaways for ML practitioners and mathematicians: models which communicate uncertainty, respond well to user corrections, are more interpretable and concise may constitute better assistants; interactive evaluation is a promising way to continually navigate the capability of these models; humans should be aware of language models' algebraic fallibility, and for that reason discern where they should be used.

Multimodal agents have demonstrated strong performance in general GUI interactions, but their application in automotive systems has been largely unexplored. In-vehicle GUIs present distinct challenges: drivers' limited attention, strict safety requirements, and complex location-based interaction patterns. To address these challenges, we introduce Automotive-ENV, the first high-fidelity benchmark and interaction environment tailored for vehicle GUIs. This platform defines 185 parameterized tasks spanning explicit control, implicit intent understanding, and safety-aware tasks, and provides structured multimodal observations with precise programmatic checks for reproducible evaluation. Building on this benchmark, we propose ASURADA, a geo-aware multimodal agent that integrates GPS-informed context to dynamically adjust actions based on location, environmental conditions, and regional driving norms. Experiments show that geo-aware information significantly improves success on safety-aware tasks, highlighting the importance of location-based context in automotive environments. We will release Automotive-ENV, complete with all tasks and benchmarking tools, to further the development of safe and adaptive in-vehicle agents.

Building open-domain conversational systems (or chatbots) that produce convincing responses is a recognized challenge. Recent state-of-the-art (SoTA) transformer-based models for the generation of natural language dialogue have demonstrated impressive performance in simulating human-like, single-turn conversations in English. This work investigates, by an empirical study, the potential for transfer learning of such models to Swedish language. DialoGPT, an English language pre-trained model, is adapted by training on three different Swedish language conversational datasets obtained from publicly available sources. Perplexity score (an automated intrinsic language model metric) and surveys by human evaluation were used to assess the performances of the fine-tuned models, with results that indicate that the capacity for transfer learning can be exploited with considerable success. Human evaluators asked to score the simulated dialogue judged over 57% of the chatbot responses to be human-like for the model trained on the largest (Swedish) dataset. We provide the demos and model checkpoints of our English and Swedish chatbots on the HuggingFace platform for public use.

Online marketplaces will be transformed by autonomous AI agents acting on behalf of consumers. Rather than humans browsing and clicking, vision-language-model (VLM) agents can parse webpages, evaluate products, and transact. This raises a fundamental question: what do AI agents buy, and why? We develop ACES, a sandbox environment that pairs a platform-agnostic VLM agent with a fully programmable mock marketplace to study this question. We first conduct basic rationality checks in the context of simple tasks, and then, by randomizing product positions, prices, ratings, reviews, sponsored tags, and platform endorsements, we obtain causal estimates of how frontier VLMs actually shop. Models show strong but heterogeneous position effects: all favor the top row, yet different models prefer different columns, undermining the assumption of a universal "top" rank. They penalize sponsored tags and reward endorsements. Sensitivities to price, ratings, and reviews are directionally human-like but vary sharply in magnitude across models. Motivated by scenarios where sellers use AI agents to optimize product listings, we show that a seller-side agent that makes minor tweaks to product descriptions, targeting AI buyer preferences, can deliver substantial market-share gains if AI-mediated shopping dominates. We also find that modal product choices can differ across models and, in some cases, demand may concentrate on a few select products, raising competition questions. Together, our results illuminate how AI agents may behave in e-commerce settings and surface concrete seller strategy, platform design, and regulatory questions in an AI-mediated ecosystem.

LLM-based coding agents repeat the same classes of mistakes across sessions because they lack a mechanism to retain corrections from human review feedback. We present a closed-loop framework in which every accepted review comment is codified as a persistent behavioral rule, progressively expanding the set of error classes the agent can self-detect. The framework combines an accumulating rule set in a version-controlled instruction file, a self-review checklist executed before code submission, and automated validation that ensures rule set integrity as it grows. In deployment across a 35+ service microservices platform, the rule set grew from 5 to 18 behavioral rules, 15+ language-specific standards, and a 15-item self-review checklist, all derived from real review feedback. We present empirical results from 11 recorded working sessions spanning code generation, PR review, incident investigation, and cross service refactoring. We observe that accumulated rules shift review effort from low-level correctness toward design-level validation, achieve a measured 0% recurrence rate for ruled-against error classes, and transfer across heterogeneous agent interfaces. We compare our approach against related work in experiential LLM learning (Reflexion, ExpeL, Voyager) and automated code review (CodeReviewer, SWE-bench agents), showing that our framework achieves persistent cross-session learning without weight updates, operates on production codebases rather than synthetic benchmarks, and addresses an orthogonal dimension (behavioral consistency over time) that existing benchmarks do not measure. The result is a coding agent that improves with every review cycle, accumulating the engineering wisdom of its human collaborators without changing a single model weight.

Large language model compliance systems are deployed on the assumption that a verdict depends on the regulatory rule it is given. We test this directly across five models and 20 regulatory and platform-policy domains: delete, swap, or negate the governing rule while holding the case fixed, and check whether the verdict changes (OCS) or the model's internal representation of compliance shifts at all (ICS-delta). Neither moves much: models' verdicts are often invariant to substantial perturbations of the supplied rule, and the guard model, evaluated here under a custom-rule adaptation of its native taxonomy, is the least rule-sensitive and least accurate of the five, barely above chance (51%, versus 90-92% for general-purpose models). This reflects easy cases more than blanket neglect: on cases where deleting the rule changes a previously correct model prediction, models do track it closely. Neither better prompting nor direct intervention on the model's internal representations closes this gap. Accuracy alone does not establish that a compliance verdict is grounded in the supplied rule.

Building generalist agents that can handle diverse tasks and evolve themselves across different environments is a long-term goal in the AI community. Large language models (LLMs) are considered a promising foundation to build such agents due to their generalized capabilities. Current approaches either have LLM-based agents imitate expert-provided trajectories step-by-step, requiring human supervision, which is hard to scale and limits environmental exploration; or they let agents explore and learn in isolated environments, resulting in specialist agents with limited generalization. In this paper, we take the first step towards building generally-capable LLM-based agents with self-evolution ability. We identify a trinity of ingredients: 1) diverse environments for agent exploration and learning, 2) a trajectory set to equip agents with basic capabilities and prior knowledge, and 3) an effective and scalable evolution method. We propose AgentGym, a new framework featuring a variety of environments and tasks for broad, real-time, uni-format, and concurrent agent exploration. AgentGym also includes a database with expanded instructions, a benchmark suite, and high-quality trajectories across environments. Next, we propose a novel method, AgentEvol, to investigate the potential of agent self-evolution beyond previously seen data across tasks and environments. Experimental results show that the evolved agents can achieve results comparable to SOTA models. We release the AgentGym suite, including the platform, dataset, benchmark, checkpoints, and algorithm implementations. The AgentGym suite is available on https://github.com/WooooDyy/AgentGym.

Ultrasound is the most widely deployed imaging modality worldwide, yet clinical AI remains fragmented into narrow single-task models that fail when device, operator, or anatomy changes. Here we present SonoCorpus, an open resource unifying 456,963 images and 1,626,085 expert masks from 53 public datasets spanning 24 clinical applications and 17 countries, and SonoBase, an interactive segmentation foundation model pretrained on it. Across fifteen evaluation datasets introducing new organs, devices, operators, and geographies, SonoBase outperforms SAM2, MedSAM2, and the concept-promptable MedSAM3 on every dataset and matches per-dataset specialist models trained on the same data; on fully external data it exceeds the accuracy these baselines achieve on their own in-distribution benchmarks. Ejection fraction derived from its segmentations falls within inter-observer variability (6.63\% error), with fewer misclassifications at the defibrillator-candidacy threshold than either promptable baseline (13\% versus 18--42\%); fetal head-circumference (1.81~mm) and gestational-age (1.2 days) errors fall below inter-observer variability. Where a baseline fails outright, one in four test cases, SonoBase recovers a usable segmentation in 81\% of them, including on handheld probes operated by minimally trained users in two low- and middle-income countries (Sierra Leone and Tanzania). Five labeled examples can help the model adapt to a new setting, and the identical training protocol transfers well to newer models such as SAM3, locating the advantage in ultrasound-specific pretraining rather than any single architecture. To ensure reproducibility and enable the community to build on SonoBase as a platform, we release all checkpoints, optimizer states, data-split indices, deduplication hashes, and starter code.

In this paper, we examine several methods of acquiring Czech data for automated fact-checking, which is a task commonly modeled as a classification of textual claim veracity w.r.t. a corpus of trusted ground truths. We attempt to collect sets of data in form of a factual claim, evidence within the ground truth corpus, and its veracity label (supported, refuted or not enough info). As a first attempt, we generate a Czech version of the large-scale FEVER dataset built on top of Wikipedia corpus. We take a hybrid approach of machine translation and document alignment; the approach and the tools we provide can be easily applied to other languages. We discuss its weaknesses and inaccuracies, propose a future approach for their cleaning and publish the 127k resulting translations, as well as a version of such dataset reliably applicable for the Natural Language Inference task - the CsFEVER-NLI. Furthermore, we collect a novel dataset of 3,097 claims, which is annotated using the corpus of 2.2M articles of Czech News Agency. We present its extended annotation methodology based on the FEVER approach, and, as the underlying corpus is kept a trade secret, we also publish a standalone version of the dataset for the task of Natural Language Inference we call CTKFactsNLI. We analyze both acquired datasets for spurious cues - annotation patterns leading to model overfitting. CTKFacts is further examined for inter-annotator agreement, thoroughly cleaned, and a typology of common annotator errors is extracted. Finally, we provide baseline models for all stages of the fact-checking pipeline and publish the NLI datasets, as well as our annotation platform and other experimental data.

Large language models can generate useful code from natural language, but their outputs come without correctness guarantees. Verifiable code generation offers a path beyond testing by requiring models to produce not only executable code, but also formal specifications and machine-checkable proofs. Progress in this direction, however, is difficult to measure: existing benchmarks are often small, focus on only one part of the pipeline, lack ground-truth proofs or rigorous specification validation, or target verification settings far from mainstream software development. We present VeriContest, a benchmark of 946 competitive-programming problems from LeetCode and Codeforces for verifiable code generation in Rust with Verus. Each problem pairs a natural language description with expert-validated formal specifications, judge-accepted Rust code, Verus-checked proofs, and positive and negative test suites. VeriContest is constructed through a three-phase pipeline that scales from manually verified seed problems to semi-automated expansion with human-in-the-loop review. To further strengthen benchmark quality, we use testing as an additional quality-assurance layer for validating postcondition completeness. VeriContest supports isolated and compositional evaluation of specification generation, code generation, proof generation, and end-to-end verified program synthesis. Evaluating ten state-of-the-art models reveals a sharp gap between coding ability and verifiable code generation: the strongest model reaches 92.18% on natural-language-to-code generation, but only 48.31% on specification generation, 13.95% on proof generation, and 5.29% end-to-end. These results identify proof and specification generation as the central bottlenecks for models and establish VeriContest as a rigorous platform for measuring and training future systems that generate code with machine-checkable correctness.

Fact checking by professionals is viewed as a vital defense in the fight against misinformation.While fact checking is important and its impact has been significant, fact checks could have limited visibility and may not reach the intended audience, such as those deeply embedded in polarized communities. Concerned citizens (i.e., the crowd), who are users of the platforms where misinformation appears, can play a crucial role in disseminating fact-checking information and in countering the spread of misinformation. To explore if this is the case, we conduct a data-driven study of misinformation on the Twitter platform, focusing on tweets related to the COVID-19 pandemic, analyzing the spread of misinformation, professional fact checks, and the crowd response to popular misleading claims about COVID-19. In this work, we curate a dataset of false claims and statements that seek to challenge or refute them. We train a classifier to create a novel dataset of 155,468 COVID-19-related tweets, containing 33,237 false claims and 33,413 refuting arguments.Our findings show that professional fact-checking tweets have limited volume and reach. In contrast, we observe that the surge in misinformation tweets results in a quick response and a corresponding increase in tweets that refute such misinformation. More importantly, we find contrasting differences in the way the crowd refutes tweets, some tweets appear to be opinions, while others contain concrete evidence, such as a link to a reputed source. Our work provides insights into how misinformation is organically countered in social platforms by some of their users and the role they play in amplifying professional fact checks.These insights could lead to development of tools and mechanisms that can empower concerned citizens in combating misinformation. The code and data can be found in http://claws.cc.gatech.edu/covid_counter_misinformation.html.

Terms of service of on-line platforms too often contain clauses that are potentially unfair to the consumer. We present an experimental study where machine learning is employed to automatically detect such potentially unfair clauses. Results show that the proposed system could provide a valuable tool for lawyers and consumers alike.

The CheckThat! lab aims to advance the development of innovative technologies designed to identify and counteract online disinformation and manipulation efforts across various languages and platforms. The first five editions focused on key tasks in the information verification pipeline, including check-worthiness, evidence retrieval and pairing, and verification. Since the 2023 edition, the lab has expanded its scope to address auxiliary tasks that support research and decision-making in verification. In the 2025 edition, the lab revisits core verification tasks while also considering auxiliary challenges. Task 1 focuses on the identification of subjectivity (a follow-up from CheckThat! 2024), Task 2 addresses claim normalization, Task 3 targets fact-checking numerical claims, and Task 4 explores scientific web discourse processing. These tasks present challenging classification and retrieval problems at both the document and span levels, including multilingual settings.

The concept of an information technology (IT) related platform is broad and covers phenomena ranging from the operating system Linux to the Internet. Such platforms are of increasing importance to innovation and value creation across many facets of industry and daily life. There is, however, a lack of common understanding in both research and industry about what is mean by the term platform when related to IT. This lack of consensus is detrimental to research and knowledge development. Thus, the aims of this study are to: (i) provide a sound definition of the IT-platform concept by identifying its distinguishing dimensions; and (ii) identify important current research directions for the IT-platform concept. To achieve these aims a systematic literature review was undertaken with 133 relevant articles taken from major information systems journals, conferences, and business publications. The study contributes by providing a sound base for future research into IT-platforms.

Recent legislation required AI platforms to provide APIs for regulators to assess their compliance with the law. Research has nevertheless shown that platforms can manipulate their API answers through fairwashing. Facing this threat for reliable auditing, this paper studies the benefits of the joint use of platform scraping and of APIs. In this setup, we elaborate on the use of scraping to detect manipulated answers: since fairwashing only manipulates API answers, exploiting scraps may reveal a manipulation. To abstract the wide range of specific API-scrap situations, we introduce a notion of proxy that captures the consistency an auditor might expect between both data sources. If the regulator has a good proxy of the consistency, then she can easily detect manipulation and even bypass the API to conduct her audit. On the other hand, without a good proxy, relying on the API is necessary, and the auditor cannot defend against fairwashing. We then simulate practical scenarios in which the auditor may mostly rely on the API to conveniently conduct the audit task, while maintaining her chances to detect a potential manipulation. To highlight the tension between the audit task and the API fairwashing detection task, we identify Pareto-optimal strategies in a practical audit scenario. We believe this research sets the stage for reliable audits in practical and manipulation-prone setups.

It is now a common business practice to buy access to large language model (LLM) inference rather than self-host, because of significant upfront hardware infrastructure and energy costs. However, as a buyer, there is no mechanism to verify the authenticity of the advertised service including the serving hardware platform, e.g. that it is actually being served using an NVIDIA H100. Furthermore, there are reports suggesting that model providers may deliver models that differ slightly from the advertised ones, often to make them run on less expensive hardware. That way, a client pays premium for a capable model access on more expensive hardware, yet ends up being served by a (potentially less capable) cheaper model on cheaper hardware. In this paper we introduce \textbf{hardware and software platform inference (HSPI)} -- a method for identifying the underlying architecture and software stack of a (black-box) machine learning model solely based on its input-output behavior. Our method leverages the inherent differences of various architectures and compilers to distinguish between different types and software stacks. By analyzing the numerical patterns in the model's outputs, we propose a classification framework capable of accurately identifying the used for model inference as well as the underlying software configuration. Our findings demonstrate the feasibility of inferring type from black-box models. We evaluate HSPI against models served on different real hardware and find that in a white-box setting we can distinguish between different s with between 83.9% and 100% accuracy. Even in a black-box setting we are able to achieve results that are up to three times higher than random guess accuracy.

We introduce {rm C{small LEVER}}, a high-quality, curated benchmark of 161 problems for end-to-end verified code generation in Lean. Each problem consists of (1) the task of generating a specification that matches a held-out ground-truth specification, and (2) the task of generating a Lean implementation that provably satisfies this specification. Unlike prior benchmarks, {rm C{small LEVER}} avoids test-case supervision, LLM-generated annotations, and specifications that leak implementation logic or allow vacuous solutions. All outputs are verified post-hoc using Lean's type checker to ensure machine-checkable correctness. We use {rm C{small LEVER}} to evaluate several few-shot and agentic approaches based on state-of-the-art language models. These methods all struggle to achieve full verification, establishing it as a challenging frontier benchmark for program synthesis and formal reasoning. Our benchmark can be found on GitHub(https://github.com/trishullab/clever) as well as HuggingFace(https://huggingface.co/datasets/amitayusht/clever). All our evaluation code is also available online(https://github.com/trishullab/clever-prover).

The emergence of "vibe coding" platforms, where users describe applications in natural language and AI agents autonomously generate full-stack software, has created a need for rigorous evaluation beyond code-level benchmarks. In order to assess them as virtual software development agencies on understanding business requirements, making architectural decisions, writing production code, handling iterative modifications, and maintaining business readiness, we introduce SWE-WebDev Bench, a 68-metric evaluation framework spanning 25 primary and 43 diagnostic metrics across seven groups, organized along three dimensions: Interaction Mode (App Creation Request (ACR) vs. App Modification Request (AMR)), Agency Angle (Product Manager (PM), Engineering, Ops), and Complexity Tier (T4 multi-role SaaS, T5 AI-native). Our evaluation (six platforms, three domains, 18 evaluation cells) reveals four recurring shortcomings in the current generation of AI app builders: (1) A specification bottleneck, where platforms compress rich business requirements into oversimplified technical plans, (2) A pervasive frontend-backend decoupling, where visually polished UIs mask absent or broken backend infrastructure, (3) A steep production-readiness cliff, where no platform scores above 60% on engineering quality and post-generation human effort varies substantially across platforms and (4) Widespread security and infrastructure failures, with no platform exceeding 65% Security Score against a 90% target and concurrency handling as low as 6%. These observations are descriptive of our sample and require larger-scale replication to establish generality. We release SWE-WebDev Bench as a community benchmark to enable such replication and help platform builders identify and address these gaps. Code and benchmark resources are available at: https://github.com/snowmountainAi/webdevbench and https://webdevbench.com/.

Finding relevant literature underpins the practice of evidence-based medicine. From 2014 to 2016, TREC conducted a clinical decision support track, wherein participants were tasked with finding articles relevant to clinical questions posed by physicians. In total, 87 teams have participated over the past three years, generating 395 runs. During this period, each team has trialled a variety of methods. While there was significant overlap in the methods employed by different teams, the results were varied. Due to the diversity of the platforms used, the results arising from the different techniques are not directly comparable, reducing the ability to build on previous work. By using a stable platform, we have been able to compare different document and query processing techniques, allowing us to experiment with different search parameters. We have used our system to reproduce leading teams runs, and compare the results obtained. By benchmarking our indexing and search techniques, we can statistically test a variety of hypotheses, paving the way for further research.

We introduce MetroLLM-Bench, a 955-case benchmark for testing language models as the policy layer of a transit kiosk. It covers six real metro systems, ranging from 37 to 414 stations, and eleven categories that include routing, fare calculation, disruptions, accessibility, and adversarial input. In each case, the model must call structured tools and submit a machine-renderable terminal state containing an outcome, a per-ticket fare quote when applicable, and a kiosk action. Fourteen deterministic scoring components form Tier 1; eight semantic-quality components form Tier 2, six of which use a language-model judge. We report Tier 1 and the combined score of both tiers. A stratified 75/25 split reserves 717 cases for training-data generation and 238 for held-out evaluation. We evaluate twenty-six models from six vendors, of which twenty-three are ranked. On the held-out partition, a 4B Qwen 3.5 student trained through parameter-efficient fine-tuning (PEFT) exceeds both GPT-5.6 tiers on Tier 1 (91.3 against 90.6 and 90.0) and matches GPT-5.4 full at maximum reasoning effort (91.4), with a 2.6 GB Q4_K_M footprint. Larger 9B and 27B students provide no further Tier 1 improvement over the 4B student at this training scale. Across the four Qwen sizes, the PEFT gain over the corresponding base model decreases from +7.03 points at 2B (three training seeds) to -0.91 at 27B; every seed shows the same direction at every size. A deterministic rule-based baseline reaches 84.6 on Tier 1, with the remaining language-model advantage concentrated in policy adaptation, compound scenarios, accessibility, and temporal reasoning. Muse Glimmer 30B leads the composite ranking, and serving configuration alone moves the Qwen 3.5-to-3.8 comparison by 2.7 Tier 1 points. The benchmark, harness, reproduction guide, and fine-tuned students are released at https://github.com/continker/metrollm-bench.

Recent agent benchmarks increasingly ground evaluation in executable environments, from code repair to web navigation, app APIs, and function calling. Yet completing consequential work beyond code requires more than producing a plausible response or valid tool call: agents must gather missing information over multiple turns, follow domain policies, coordinate dependent tools, and realize the correct persistent state transition without collateral effects. In this paper, we introduce Thinkingbox, a sandbox for tool-agent-user interaction that provides isolated MCP-compatible tool sessions, complete execution traces, and outcome evaluation over terminal backend state. Built on this sandbox, Thinkingbox-bench contains 507 policy-conditioned workflows across numerous scenarios, including retail, hospitality, auto insurance, neobank internal IT, and consulting IT/HR support. Each attempt is evaluated by task-specific executable checks that accept valid trajectories while rejecting wrong, missing, or extra effects; designated tasks additionally check required properties of the final response. Across proprietary and open-weight models, the strongest achieves 65.36% pass@1, but only 25.25% pass^20. Moreover, many failed trials show clean termination and valid state-changing actions, showing that response or tool-call-level signals are not clear proxies for end-to-end task completion. Thinkingbox-bench reveals a large gap between occasionally finding a successful trajectory and reliably completing stateful business tasks. We release both Thinkingbox and Thinkingbox-Bench: https://github.com/microsoft/thinkingbox

Software is used in critical applications in our day-to-day life and it is important to ensure its correctness. One popular approach to assess correctness is to evaluate software on tests. If a test fails, it indicates a fault in the software under test; if all tests pass correctly, one may assume that the software is correct. However, the reliability of these results depends on the test suite considered, and there is a risk of false negatives (i.e. software that passes all available tests but contains bugs because some cases are not tested). Therefore, it is important to consider error-inducing test cases when evaluating software. To support data-driven creation of such a test-suite, which is especially of interest for testing software synthesized from large language models, we curate a dataset (Codehacks) of programming problems together with corresponding error-inducing test cases (i.e., "hacks"). This dataset is collected from the wild, in particular, from the Codeforces online judge platform. The dataset comprises 288,617 hacks for 5,578 programming problems, each with a natural language description, as well as the source code for 2,196 submitted solutions to these problems that can be broken with their corresponding hacks. Keywords: competitive programming, language model, dataset

Evolving documents, such as laws, tax codes, and software documentation, are amended, replaced, and sometimes reverted over time, so a question has different correct answers at different dates. In contrast to encyclopedic knowledge, where an old fact is simply overwritten, an amendment is itself an official text that states what it replaces and when it takes effect, and the earlier version stays correct for its validity period. The central challenge is therefore version resolution, that is, identifying the version in force on the queried date. Existing temporal QA datasets treat time only as an annotation, so version resolution stays untested. We present TIDE, an expert-verified benchmark of 3,050 QA pairs over 644 official customs instruments issued between 1969 and 2025 by the Government of Bangladesh, covering eight task types over deeply code-mixed documents that are heterogeneous in layout and dated in two calendars. In addition, we evaluate nine recent LLMs under a single protocol across parametric, gold-context, and retrieval access, scored by a three-judge LLM council with a hard date gate separating correct meaning from correct time. The best macro-averaged accuracy is only 68.5%. Resolving a version from an implicit date reaches 59.7%, and detecting that the supplied version does not govern the query reaches only 26.7%. Models are more likely to find correct versions than to reject incorrect ones, and they tend to follow a confident parametric answer over the supplied authoritative text. All code and data are available at https://github.com/icsetepa44/TIDE

Agent benchmarks often evaluate only final answers even when agents run on stateful runtimes. We argue this under-specifies what is being evaluated: the proper unit is a declared model-plus-runtime configuration whose failures can occur in evidence acquisition, runtime routing, safety boundaries, or repeated execution. We present ClawProBench, a trace-aware benchmark for runtime-native agent evaluation instantiated on OpenClaw, a live agent runtime with workspace tools and native surfaces for browsing, memory, messaging, scheduling, skills, and subagents. ClawProBench defines two tracks: a 102-scenario full profile with live workspace and native-runtime routing tasks, and a frozen 68-scenario holdout with closed-world JSON output contracts for robust ranking. Trials are scored from execution traces via a safety-gated formula combining correctness, process quality, and efficiency, preserving failure evidence for audit. Our anonymous artifact includes benchmark definitions, scoring code, manifests and sanitized traces. We evaluate 68 configurations on the full profile and 37 on holdout. The top safety-gated average trace score is 0.7671. Native-runtime tasks underperform workspace-live tasks (0.5238 vs. 0.6415). On holdout, pass@k-any outperforms strict three-trial pass (0.6638 vs. 0.2890), while full-profile and holdout rankings show weak alignment (Spearman 0.1300). Rankings based purely on correctness differ substantially from process-aware, safety-gated and strict-pass views. Final-answer leaderboards may hide native-surface weaknesses, one-off successes and trace-local agent failure modes.

The proliferation of harmful content on online platforms is a major societal problem, which comes in many different forms including hate speech, offensive language, bullying and harassment, misinformation, spam, violence, graphic content, sexual abuse, self harm, and many other. Online platforms seek to moderate such content to limit societal harm, to comply with legislation, and to create a more inclusive environment for their users. Researchers have developed different methods for automatically detecting harmful content, often focusing on specific sub-problems or on narrow communities, as what is considered harmful often depends on the platform and on the context. We argue that there is currently a dichotomy between what types of harmful content online platforms seek to curb, and what research efforts there are to automatically detect such content. We thus survey existing methods as well as content moderation policies by online platforms in this light and we suggest directions for future work.

Serving an answer from a large language model requires deciding when to abstain, yet a verifier's ranking accuracy alone does not determine the error rate among served answers. We introduce PriceCheck, which builds a compact family of decision rules from label-free checks such as re-solving a problem. Each check has a price: its agreement rates on correct and incorrect answers and its cost per run. Prices fitted on a small, class-enriched labelled set compose into predictions of a schedule's coverage and cost, guiding which checks to run and when to stop. A calibration test then selects a schedule at a stated selective-risk target. In mathematics, the selected schedules serve 76.1% of answers on average and keep held-out selective risk below 1.5% on all 15 splits. Under the shared testing protocol, PriceCheck serves more answers at that target than reward models, a prompted judge, the generator's confidence and a trained correctness classifier. At matched coverage, it keeps the fewest wrong answers among these scorers. Across 118 diagnostic schedules, price-based coverage predictions have a rank correlation of 0.97 with observed coverage. These results show that choosing how checks are combined and stopped matters alongside how well a verifier ranks answers. Code is available at https://github.com/js-lee-AI/PriceCheck.

Computer-using agents drive real software through the screen -- clicking and typing -- but they solve every task from scratch: asked to repeat a task, an agent re-reads the screen, re-reasons every tap, and pays the full cost again. We present PreAct, which lets such an agent get faster on tasks it has done before. The first time it succeeds, PreAct compiles the run into a small state-machine program-states that check the screen, transitions that act-and on later runs replays it directly instead of invoking the agent 8.5-13x faster, with no per-step language-model calls. Replay is not blind: at each step PreAct checks that the screen matches what the program expects before acting, and hands control back to the agent the moment something is off. PreAct applies the same discipline when deciding what to keep: a freshly compiled program enters the store only if, re-run from a clean state, an independent evaluator confirms it solved the task-catching programs that replay to their last step yet leave the task undone. Across a mobile, a desktop, and a web benchmark, this store-time check separates repeated runs that improve from ones that degrade as faulty programs accumulate, worth 1.75-2.6 tasks per benchmark, the same direction on all three; a fallback that explores afresh when no program fits brings PreAct level with a strong record-and-replay baseline. We also report what did not matter: prompt wording, runtime guardrails, and whether a language model or a plain embedding retriever selects which program to reuse.

In this work, we propose GUIPilot, an approach for detecting inconsistencies between the mobile design and their implementations. The mobile design usually consists of design mock-ups that specify (1) the expected screen appearances (e.g., widget layouts, colors, and shapes) and (2) the expected screen behaviors, regarding how one screen can transition into another (e.g., labeled widgets with textual description). Given a design mock-up and the implementation of its application, GUIPilot reports both their screen inconsistencies as well as process inconsistencies. On the one hand, GUIPilot detects the screen inconsistencies by abstracting every screen into a widget container where each widget is represented by its position, width, height, and type. By defining the partial order of widgets and the costs of replacing, inserting, and deleting widgets in a screen, we convert the screen-matching problem into an optimizable widget alignment problem. On the other hand, we translate the specified GUI transition into stepwise actions on the mobile screen (e.g., click, long-press, input text on some widgets). To this end, we propose a visual prompt for the vision-language model to infer widget-specific actions on the screen. By this means, we can validate the presence or absence of expected transitions in the implementation. Our extensive experiments on 80 mobile applications and 160 design mock-ups show that (1) GUIPilot can achieve 94.5% precision and 99.6% recall in detecting screen inconsistencies, outperforming the state-of-the-art approach, such as GVT, by 66.2% and 56.6% respectively, and (2) GUIPilot reports zero errors in detecting process inconsistencies. Furthermore, our industrial case study on applying GUIPilot on a trading mobile application shows that GUIPilot has detected nine application bugs, and all the bugs were confirmed by the original application experts.

AI-powered people search platforms are increasingly used in recruiting, sales prospecting, and professional networking, yet no widely accepted benchmark exists for evaluating their performance. We introduce PeopleSearchBench, an open-source benchmark that compares four people search platforms on 119 real-world queries across four use cases: corporate recruiting, B2B sales prospecting, expert search with deterministic answers, and influencer/KOL discovery. A key contribution is Criteria-Grounded Verification, a factual relevance pipeline that extracts explicit, verifiable criteria from each query and uses live web search to determine whether returned people satisfy them. This produces binary relevance judgments grounded in factual verification rather than subjective holistic LLM-as-judge scores. We evaluate systems on three dimensions: Relevance Precision (padded nDCG@10), Effective Coverage (task completion and qualified result yield), and Information Utility (profile completeness and usefulness), averaged equally into an overall score. Lessie, a specialized AI people search agent, performs best overall, scoring 65.2, 18.5% higher than the second-ranked system, and is the only system to achieve 100% task completion across all 119 queries. We also report confidence intervals, human validation of the verification pipeline (Cohen's kappa = 0.84), ablations, and full documentation of queries, prompts, and normalization procedures. Code, query definitions, and aggregated results are available on GitHub.

Spec-driven development (SDD) with AI coding agents provides a structured workflow, but agents often remain "context blind" in large, evolving repositories, leading to hallucinated APIs and architectural violations. We present Spec Kit Agents, a multi-agent SDD pipeline (with PM and developer roles) that adds phase-level, context-grounding hooks. Read-only probing hooks ground each stage (Specify, Plan, Tasks, Implement) in repository evidence, while validation hooks check intermediate artifacts against the environment. We evaluate 128 runs covering 32 features across five repositories. Context-grounding hooks improve judged quality by +0.15 on a 1-5 composite LLM-as-judge score (+3.0 percent of the full score; Wilcoxon signed-rank, p < 0.05) while maintaining 99.7-100 percent repository-level test compatibility. We further evaluate the framework on SWE-bench Lite, where augmentation hooks improve baseline by 1.7 percent, achieving 58.2 percent Pass@1.

Large language models (LLMs) have made misinformation inexpensive to produce but not to verify, creating a growing asymmetry in the information ecosystem. Under tight time, labor, and budget constraints, media organizations, platforms, and fact-checkers rely on screening to prioritize which content to verify. We introduce VEX-Bench, a unified benchmark for evaluating the verification complexity of LLM-generated misinformation, as perceived during screening, across models and generation methods. Verification complexity is assessed along multiple dimensions derived from journalistic and fact-checking practices, capturing checkability, harm potential, source credibility signals, imposter legitimacy, and expected verification effort. We define the VEX score as an integrated measure combining elicitation yield and verification complexity to quantify how generated content consumes limited verification capacity. We construct a benchmark spanning two misinformation categories, 6 high-stakes domains, and 60 real-world topics, and evaluate 7 frontier LLMs and 7 generation methods, yielding 5{,}880 articles. We employ an LLM-as-judge for scalable evaluation and validate it using content-analysis methodology, including ordinal Krippendorff α for inter-annotator reliability, complemented by fact-checking agents for verification. Our findings show that no single method dominates all dimensions, underscoring the need for multi-dimensional evaluation. LLMs can generate high-VEX misinformation at 3times to 169times lower cost than agent-based verification. Such content is often prioritized during screening, consuming scarce verification resources and introducing a systematic risk of misallocation in resource-constrained verification systems. The code is publicly available in our https://github.com/HanxunH/VEX-Bench{GitHub repository}.

Open-ended language-model benchmarks usually inherit a judge: a human preference panel, another model, or a brittle exact-match key. We introduce FlavourBench, an automated benchmark in which a versioned culinary system supplies dense, executable ground truth. Each task presents eight ingredients and asks for a three-ingredient portfolio; before model execution, Epicure scores all 56 possible portfolios. We evaluate 27 frontier endpoints on an identical 534-task core spanning substitution, pairing, and constrained composition. Every ranked model has exactly 89 valid responses per panel and family (14,418 model-task cells total), eliminating differential missingness from the leaderboard. The FlavourBench Score is the equal-family mean of the frozen task scores. We use 50,000 anchor-cluster bootstrap replicates for simultaneous 95% score bands and 100,000 sign-flip draws for all 351 paired model contrasts, with Holm control. The two independently compiled panels correlate at r = 0.89 (rank rho = 0.80). Grok 4.6 has the largest point estimate at 65.1 (simultaneous 95% CI 61.0-69.2); 101 of 351 model pairs are resolved. The release includes the prompts, all portfolio score maps, raw responses, exact routes, content hashes, and an offline verifier that reconstructs every result.

Benchmarks are essential for consistent evaluation and reproducibility. The integration of Artificial Intelligence into Software Engineering (AI4SE) has given rise to numerous benchmarks for tasks such as code generation and bug fixing. However, this surge presents challenges: (1) scattered benchmark knowledge across tasks, (2) difficulty in selecting relevant benchmarks, (3) the absence of a uniform standard for benchmark development, and (4) limitations of existing benchmarks. In this paper, we review 173 studies and identify 204 AI4SE benchmarks. We classify these benchmarks, analyze their limitations, and expose gaps in practices. Based on our review, we created BenchScout, a semantic search tool to find relevant benchmarks, using automated clustering of the contexts from associated studies. We conducted a user study with 22 participants to evaluate BenchScout's usability, effectiveness, and intuitiveness which resulted in average scores of 4.5, 4.0, and 4.1 out of 5. To advance benchmarking standards, we propose BenchFrame, a unified method to enhance benchmark quality. As a case study, we applied BenchFrame to the HumanEval benchmark and addressed its main limitations. This led to HumanEvalNext, featuring (1) corrected errors, (2) improved language conversion, (3) expanded test coverage, and (4) increased difficulty. We then evaluated ten state-of-the-art code language models on HumanEval, HumanEvalPlus, and HumanEvalNext. On HumanEvalNext, models showed a pass@1 score reduction of 31.22% and 19.94% compared to HumanEval and HumanEvalPlus, respectively.

We introduce a novel framework named ClarifyGPT, which aims to enhance code generation by empowering LLMs with the ability to identify ambiguous requirements and ask targeted clarifying questions. In particular, ClarifyGPT first detects whether a given requirement is ambiguous by performing a code consistency check. If it is ambiguous, ClarifyGPT prompts an LLM to generate targeted clarifying questions. After receiving question responses, ClarifyGPT refines the ambiguous requirement and inputs it into the same LLM to generate a final code solution. To evaluate our ClarifyGPT, we first conduct a human evaluation involving ten participants who use ClarifyGPT for code generation on two publicly available benchmarks: MBPP-sanitized and MBPP-ET. The results show that ClarifyGPT elevates the performance (Pass@1) of GPT-4 from 70.96% to 80.80% on MBPP-sanitized. Furthermore, to perform large-scale automated evaluations of ClarifyGPT across different LLMs and benchmarks without requiring user participation, we introduce a high-fidelity simulation method to simulate user responses. The automated evaluation results also demonstrate that ClarifyGPT can significantly enhance code generation performance compared to the baselines. In particular, ClarifyGPT improves the average performance of GPT-4 and ChatGPT across four benchmarks from 68.02% to 75.75% and from 58.55% to 67.22%, respectively. We believe that ClarifyGPT can effectively facilitate the practical application of LLMs in real-world development environments.

Data visualizations are common in the real-world. We often use them in data sources such as scientific documents, news articles, textbooks, and social media to summarize key information in a visual form. Charts can also mislead its audience by communicating false information or biasing them towards a specific agenda. Verifying claims against charts is not a straightforward process. It requires analyzing both the text and visual components of the chart, considering characteristics such as colors, positions, and orientations. Moreover, to determine if a claim is supported by the chart content often requires different types of reasoning. To address this challenge, we introduce ChartCheck, a novel dataset for fact-checking against chart images. ChartCheck is the first large-scale dataset with 1.7k real-world charts and 10.5k human-written claims and explanations. We evaluated the dataset on state-of-the-art models and achieved an accuracy of 73.9 in the finetuned setting. Additionally, we identified chart characteristics and reasoning types that challenge the models.

Programmable logic controllers (PLCs) run industrial plants, and large language models can already generate independent program organization units (POUs) for them. Whether such logic integrates into an existing PLC project and then runs correctly has been checked only in limited tests. We present SemaPLC, a project-grounded and verification-gated agent harness assembled from conventional tools but governed by a strict completion rule. Rather than stopping when the model judges its own output adequate, SemaPLC declares a task complete only when logged external checks confirm it. Those checks cover the specification, the compilation, and the behavior on a live runtime. On 117 independent-POU tasks matching existing benchmarks, it attains the highest strict verified pass rate on all seven models (72.6\% mean). On a project-context track of 65 tasks whose generated logic must compile and run inside a real project, it attains the highest mean on integrated compilation, static behavior, and dynamic behavior. Of the three layers, dynamic behavior is the most revealing. We measure it by deploying the generated and the reference logic to a live PLC runtime and comparing their executed traces. All methods fall within 10 static points of one another, whereas dynamic scores separate them sharply, from 22.4 to 31.4 for the baselines against 52.2 for SemaPLC. Overall, our verification-gated harness raises the mean at every layer and most sharply at runtime. Execution, not static scoring, is the faithful test of whether generated control logic actually works. SemaPLC is open-sourced at https://github.com/midea-ai/SemaPLC.

Autonomous software generation (ASG) aims to turn human requirements into executable applications, but delivering these applications does not necessarily establish that their interacting components satisfy the specified behavioral requirements. We introduce GameASG-Bench, a benchmark that makes behavioral testability part of the generation task for game development. Our design declares an evaluation interface specification before generation, fixing legal starting scenarios, player-level actions, stable snapshots, rejection behavior, and invariants while leaving private implementations open. Concretely, we include: (i) static L1 checks that assess source-level compliance; and (ii) browser-executed L2 checks that combine semantic observations with real input and runtime evidence. We implement this protocol as 47 browser-native game-generation tasks spanning 12 primary genres and both 2D and 3D interaction, each with executable checks and an independently verified reference implementation. Our experiments answer four key questions about end-to-end agent performance, tool access and nominal turn budget, reasoning effort, and harness choice. Across nine agent stacks, the highest observed mean L2 check pass rate is 93.2%, yet the highest observed strict task success rate, requiring all L1 and applicable L2 prerequisite and core requirement checks, is only 55.3% (26/47 tasks). For DeepSeek-V4-Flash, full tool access and larger nominal turn budgets yield more strict task successes, while the strict task success rate is not monotonic in reasoning effort. Both tested harnesses achieve 18 strict task successes, but only ten tasks succeed under both. These results expose task-level compliance gaps that high average check pass rates actually obscure.

We introduce PTXBench, a benchmark for evaluating and adapting large language models (LLMs) to use architecture-specific PTX for GPU kernel optimization. PTXBench measures functional correctness, whether selected target instructions execute at runtime, and speedup over frontier libraries across GEMM and attention workloads on H100 and B200 GPUs. Our evaluation shows that architecture-specific PTX capability remains uneven: success rates fall substantially on complex attention backward workloads, and executing the target instructions does not necessarily translate into competitive performance. No evaluated model consistently matches frontier libraries across the suite. We further adapt Qwen3.6-27B using supervised fine-tuning. Repair-conditioned training improves several tasks, but generalization remains uneven; data coverage, balance, and the quality of the reasoning teacher matter in addition to dataset size. PTXBench provides an auditable testbed for measuring and improving LLMs' ability to exploit evolving GPU architectures.