Daily Papers

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.

Real-world video benchmarks provide broad coverage, but their fixed clips entangle event count, rate, duration, and visual complexity, making failure modes hard to isolate. While existing programmatic benchmarks offer better control, they score only the final answer rather than auditing reported events against executable ground truth. To bridge this gap, we introduce trace-grounded parametric profiling for event counting in three controlled video tasks: bouncing-ball wall contacts, visual blinks, and categorical state transitions. Across 2,190 videos, we vary event count N and frequency F while holding rendering fixed. Each video includes an executable event trace for capability-surface estimation and timestamp-level evaluation. Our results reveal a staged temporal failure. At an 80% reliability threshold, Gemini 3.6 Flash reliably counts persistent state transitions up to 12 events at 0.5 and 1.0 Hz, yet demonstrates no reliable positive-count region for transient blinking events. Thus, event representation dictates whether a model initially accesses evidence -- a limitation that compounds as count and frequency increase. In the high-count, high-frequency regime, only 0.2% of final counts are correct and the model recovers just 18.1% of true events. To test if visual access is the primary bottleneck, we increase sampling rate. Although this boosts Bounce Ball accuracy from 19.6% to 29.3%, the reported sequence agrees with ground truth only 3.7% of the time. Extra frames can therefore inflate final scores without producing faithful event recovery. Different prompting strategies yield similarly limited gains, and real-world video evaluations show the same concentration of success at low event counts. Ultimately, trace-grounded profiling shifts video evaluation from aggregate accuracy metrics to a detailed diagnostic of where temporal reasoning fails.

Large language model-powered coding agents are increasingly used to modify existing code repositories, for example, by adding features or fixing bugs. Yet existing repository-level benchmarks typically evaluate only whether the final patch passes tests. Satisfying a user request requires a long chain of interdependent reasoning and decisions: an agent must recover explicit and implicit requirements, formulate a repository-grounded implementation plan, and translate it into correct code. A pass/fail outcome cannot characterize how an unsuccessful trajectory diverges from the requirements and implementation process needed for a correct patch. To address this gap, we introduce SWE-RPG, a repository-level benchmark that combines executable patch evaluation with validated ground-truth references (GTs) for (1) Requirement Clarification and (2) Implementation Planning. These intermediate GTs support retrospective, GT-aligned diagnosis of complete coding-agent trajectories across clarification, planning, code generation, and artifact submission. SWE-RPG comprises 163 tasks from 31 Python and Java repositories, including 113 bug fixes and 50 feature additions. We evaluate 3 coding agents, including Claude Code, Codex, and OpenCode, with 6 large language model backends, including Claude-Sonnet-5 and GPT-5.6-Terra. Results show that the evaluated popular coding agents still struggle to implement user requests in existing repositories, achieving an average resolved rate of only 31.5% on SWE-RPG. Intermediate-GT diagnosis further identifies implicit requirement recovery as the main bottleneck, accounting for 24.5%--46.0% of agent runs. This result suggests implicit-requirement recovery as a key candidate direction for improving coding agents. The benchmark data and evaluation code are available at https://github.com/Xin-Zhou-smu/SWE-RPG-Bench.

Recent RL methods have substantially improved the reasoning abilities of LLMs. Existing reward designs mainly follow two paradigms: (1) Reinforcement learning with verifiable rewards (RLVR) derives outcome signals from executable checks or ground-truth answers, but provides limited guidance for intermediate reasoning behaviors. (2) Rubrics-as-reward (RaR) goes beyond final-answer checking by using natural-language rubrics to assess reasoning quality and task compliance, but often requires instance-specific rubrics and substantial design effort. To address these issues, we introduce Metacognition-as-Reward (MaR), a metacognition-inspired RL framework that guides LLM reasoning through two general process dimensions: i) metacognitive knowledge, which identifies task-relevant information without hand-crafted instance-specific rubrics, and ii) metacognitive regulation, which plans and adjusts the reasoning process to provide reward guidance beyond final-answer outcomes. MaR scaffolds model rollouts into explicit metacognitive components and optimizes them with a trajectory-level reward over task knowledge coverage, regulation fidelity, and final-answer correctness. In this way, MaR extends reward feedback to reasoning trajectories while grounding the reward signals in general metacognitive dimensions. Experiments on 22 benchmarks show that MaR consistently improves model performance, achieving up to a 7.7% gain over the base model and up to an 11.0% gain over vanilla DAPO. Notably, Qwen3.5-9B + MaR narrows the gap to frontier models, surpassing GPT-OSS-120B on overall average and outperforming stronger models on several individual benchmarks. Process-level analysis further shows substantial improvements in reasoning process quality. MaR also generalizes to out-of-domain datasets, where MaR-trained models improve over their corresponding base models on average.

When MLLMs fail at Science, Technology, Engineering, and Mathematics (STEM) visual reasoning, a fundamental question arises: is it due to perceptual deficiencies or reasoning limitations? Through systematic scaling analysis that independently scales perception and reasoning components, we uncover a critical insight: scaling perception consistently outperforms scaling reasoning. This reveals perception as the true lever limiting current STEM visual reasoning. Motivated by this insight, our work focuses on systematically enhancing the perception capabilities of MLLMs by establishing code as a powerful perceptual medium--executable code provides precise semantics that naturally align with the structured nature of STEM visuals. Specifically, we construct ICC-1M, a large-scale dataset comprising 1M Image-Caption-Code triplets that materializes this code-as-perception paradigm through two complementary approaches: (1) Code-Grounded Caption Generation treats executable code as ground truth for image captions, eliminating the hallucinations inherent in existing knowledge distillation methods; (2) STEM Image-to-Code Translation prompts models to generate reconstruction code, mitigating the ambiguity of natural language for perception enhancement. To validate this paradigm, we further introduce STEM2Code-Eval, a novel benchmark that directly evaluates visual perception in STEM domains. Unlike existing work relying on problem-solving accuracy as a proxy that only measures problem-relevant understanding, our benchmark requires comprehensive visual comprehension through executable code generation for image reconstruction, providing deterministic and verifiable assessment. Code is available at https://github.com/TongkunGuan/Qwen-CodePercept.

LLM-based agents struggle to execute complex, multi-step Standard Operating Procedures (SOPs) that are fundamental to industrial automation. Existing benchmarks fail to capture the procedural complexity and tool orchestration demands of real-world workflows. We introduce SOP-Bench, a benchmark of 2,000+ tasks from human expert-authored SOPs across 12 business domains (healthcare, logistics, finance, content moderation, etc.). Using a human-AI collaborative framework, experts crafted authentic SOPs while AI generated artifacts (tools, APIs, datasets), all human-validated, yielding realistic tasks with executable interfaces and ground-truth outputs. SOP-Bench serves as a research enabler for systematically investigating agent architectures, model capabilities, and deployment considerations across diverse procedural tasks. We demonstrate its utility through illustrative experiments with a subset of frontier models across Function-Calling (FC) and ReAct agents, revealing critical insights. For example, (1) newer models do not guarantee better performance - Claude 4 family outperforms Claude 4.5 family on ReAct tasks (Claude 4 Opus: 72.4% vs. Claude 4.5 Sonnet: 63.3% task success rate), demonstrating that production upgrades require validation; (2) no single model-agent combination dominates: best performances range from 57% to 100% depending on domain. These examples illustrate how SOP-Bench enables isolating and studying specific dimensions of agent performance without costly production experiments. Our goal is not to rank model capabilities or build optimal agents, but to provide a rigorous evaluation framework that enables the researchers and practitioners to systematically investigate agent design choices, model selection, and deployment strategies. We release the benchmark at https://github.com/amazon-science/sop-bench.

Self-evolution of multimodal large language models (MLLMs) remains a critical challenge: pseudo-label-based methods suffer from progressive quality degradation as model predictions drift, while template-based methods are confined to a static set of transformations that cannot adapt in difficulty or diversity. We contend that robust, continuous self-improvement requires not only deterministic external feedback independent of the model's internal certainty, but also a mechanism to perpetually diversify the training distribution. To this end, we introduce EVE (Executable Visual transformation-based self-Evolution), a novel framework that entirely bypasses pseudo-labels by harnessing executable visual transformations continuously enriched in both variety and complexity. EVE adopts a Challenger-Solver dual-policy architecture. The Challenger maintains and progressively expands a queue of visual transformation code examples, from which it synthesizes novel Python scripts to perform dynamic visual transformations. Executing these scripts yields VQA problems with absolute, execution-verified ground-truth answers, eliminating any reliance on model-generated supervision. A multi-dimensional reward system integrating semantic diversity and dynamic difficulty calibration steers the Challenger to enrich its code example queue while posing progressively more challenging tasks, preventing mode collapse and fostering reciprocal co-evolution between the two policies. Extensive experiments demonstrate that EVE consistently surpasses existing self-evolution methods, establishing a robust and scalable paradigm for verifiable MLLM self-evolution. The code is available at https://github.com/0001Henry/EVE .

We introduce SLR, an end-to-end framework for systematic evaluation and training of Large Language Models (LLMs) via Scalable Logical Reasoning. Given a user's task specification, SLR enables scalable, automated synthesis of inductive reasoning tasks with precisely controlled difficulty. For each task, SLR synthesizes (i) a latent ground-truth rule, (ii) an executable validation program used by a symbolic judge to deterministically verify model outputs, and (iii) an instruction prompt for the reasoning task. Using SLR, we create SLR-Bench, a benchmark comprising over 19k prompts spanning 20 curriculum levels that progressively increase in relational, arithmetic, and recursive complexity. Large-scale evaluation reveals that contemporary LLMs readily produce syntactically valid rules, yet often fail at correct logical inference. Recent reasoning LLMs do somewhat better, but incur substantial increases in test-time compute, sometimes exceeding 15k completion tokens. Finally, logic-tuning via SLR doubles Llama-3-8B accuracy on SLR-Bench, achieving parity with Gemini-Flash-Thinking at a fraction of computational cost. SLR is fully automated, requires no human annotation, ensures dataset novelty, and offers a scalable environment for probing and advancing LLMs' reasoning capabilities.

We introduce MedAgentGYM, the first publicly available training environment designed to enhance coding-based medical reasoning capabilities in large language model (LLM) agents. MedAgentGYM comprises 72,413 task instances across 129 categories derived from authentic real-world biomedical scenarios. Tasks are encapsulated within executable coding environments, each featuring detailed task descriptions, interactive feedback mechanisms, verifiable ground-truth annotations, and scalable training trajectory generation. Extensive benchmarking of over 30 LLMs reveals a notable performance disparity between commercial API-based models and open-source counterparts. Leveraging MedAgentGYM, Med-Copilot-7B achieves substantial performance gains through supervised fine-tuning (+36.44%) and continued reinforcement learning (+42.47%), emerging as an affordable and privacy-preserving alternative competitive with gpt-4o. By offering both a comprehensive benchmark and accessible, expandable training resources within unified execution environments, MedAgentGYM delivers an integrated platform to develop LLM-based coding assistants for advanced biomedical research and practice.

AI-agent guardrails are memoryless: each message is judged in isolation, so an adversary who spreads a single attack across dozens of sessions slips past every session-bound detector because only the aggregate carries the payload. We make three contributions to cross-session threat detection. (1) Dataset. CSTM-Bench is 26 executable attack taxonomies classified by kill-chain stage and cross-session operation (accumulate, compose, launder, inject_on_reader), each bound to one of seven identity anchors that ground-truth "violation" as a policy predicate, plus matched Benign-pristine and Benign-hard confounders. Released on Hugging Face as intrinsec-ai/cstm-bench with two 54-scenario splits: dilution (compositional) and cross_session (12 isolation-invisible scenarios produced by a closed-loop rewriter that softens surface phrasing while preserving cross-session artefacts). (2) Measurement. Framing cross-session detection as an information bottleneck to a downstream correlator LLM, we find that a session-bound judge and a Full-Log Correlator concatenating every prompt into one long-context call both lose roughly half their attack recall moving from dilution to cross_session, well inside any frontier context window. Scope: 54 scenarios per shard, one correlator family (Anthropic Claude), no prompt optimisation; we release it to motivate larger, multi-provider datasets. (3) Algorithm and metric. A bounded-memory Coreset Memory Reader retaining highest-signal fragments at K=50 is the only reader whose recall survives both shards. Because ranker reshuffles break KV-cache prefix reuse, we promote CSR_prefix (ordered prefix stability, LLM-free) to a first-class metric and fuse it with detection into CSTM = 0.7 F_1(CSDA@action, precision) + 0.3 CSR_prefix, benchmarking rankers on a single Pareto of recall versus serving stability.

As production code evolves, the test suite must co-evolve to remain effective. Existing benchmarks for test evolution operate at method-level granularity with pre-paired inputs, bypassing the task of locating affected tests from the full project and excluding the need for new tests entirely. We present TEBench, the first project-level benchmark for test evolution. Given a project repository and a code-changing commit, TEBench requires systems to autonomously identify tests requiring modification, determine where new tests are needed, and produce the corresponding test patch. We construct TEBench through a four-stage pipeline over Defects4J projects, curating 314 task instances from 10 projects with developer-written ground truth. Each instance is annotated with one or more of three evolution types: Test-Breaking (tests that fail), Test-Stale (tests that pass but no longer meaningfully validate updated behavior), and Test-Missing (new tests needed for introduced behavior). We evaluate seven configurations spanning three industrial agent frameworks (Claude Code, Codex CLI, OpenCode) and six base models, alongside a heuristic baseline. All seven configurations converge on an identification F1 of 45.7% to 49.4%, revealing a shared performance ceiling across both frameworks and base models. Test-Stale is the most challenging type, averaging F1 around 36%, since configurations rely on execution failure signals and lack proactive semantic reasoning. On the update task, configurations produce highly executable test modifications whose surface form diverges substantially from ground truth. Trajectory analysis reveals a reactive "execute-fail-fix" loop that succeeds for breaking tests but structurally cannot address stale or missing tests. TEBench is available at https://github.com/iSEngLab/TEBench with a leaderboard at https://tebench-leadership.vercel.app.

Large Language Models (LLMs) are evolving from text generators into reasoning agents. This transition makes their ability to use external tools a critical capability. However, evaluating this skill presents a significant challenge. Existing benchmarks are often limited by their reliance on synthetic tools and severely constrained action spaces. To address these limitations, we introduce MCPVerse, an expansive, real-world benchmark for evaluating agentic tool use. MCPVerse integrates more than 550 real-world, executable tools to create an unprecedented action space exceeding 140k tokens, and employs outcome-based evaluation with real-time ground truth for time-sensitive tasks. We benchmarked the state-of-the-art LLMs across three modes (Oracle, Standard, and Max-Scale), revealing that while most models suffer performance degradation when confronted with larger tool sets, the agentic models, such as Claude-4-Sonnet, can effectively leverage expanded exploration spaces to improve accuracy. This finding not only exposes the limitations of state-of-the-art models in complex, real-world scenarios but also establishes MCPVerse as a critical benchmark for measuring and advancing agentic tool use capabilities.

Spatial reasoning remains a persistent challenge for multimodal large language models (MLLMs). Existing approaches largely rely on large-scale, statically curated datasets, where all training samples are treated uniformly regardless of the model's evolving capabilities. This static paradigm is inherently data-inefficient: training capacity is often spent on samples that are either trivial or overly difficult for the model at its current stage. To address this limitation, we propose Ouroboros-Spatial, a self-evolving training framework in which the model plays dual roles as a proposer and a solver. In each iteration, a frozen proposer generates spatial question-answer (QA) pairs from 3D scene metadata and raw video frames, together with executable code for deriving reliable ground truth. A learnable solver is then fine-tuned on the accepted samples, and its per-sample prediction confidence is used as a difficulty signal. This signal is fed back to the proposer in the next iteration, guiding it to generate questions better matched to the solver's current capabilities. Through this closed-loop design, the training distribution co-evolves with model ability, reducing redundant trivial examples while filtering out ambiguous or uninformative samples with limited learning value. Across six spatial reasoning benchmarks, Ouroboros-Spatial substantially improves Qwen3-VL-4B and Qwen3-VL-8B while using an order of magnitude fewer training examples than recent large-scale curated datasets. On VSI-Bench, it yields absolute gains of 9.9 and 6.8 points for the 4B and 8B models, respectively, enabling both to outperform a wide range of strong open-source and proprietary baselines.

Multi-scene navigation (clearing an objective in one bounded space and then crossing a portal into the next) is a defining feature of contemporary 3D games, but authoring it is laborious: every portal must have consistent endpoints on both sides, each interior must remain navigable once it is furnished, and the resulting connectivity must be kept consistent across many files. Recent large language model (LLM) and multimodal LLM (MLLM) scene generators have made single-interior synthesis dramatically cheaper, yet they produce one scene at a time and cannot, by naive repetition, yield a connected multi-scene world. We identify three obstacles that single-scene methods leave unsolved: cross-scene consistency, in-scene navigability, and the evaluation of whether a transition actually works. We present MAGIC, a prompt-to-project system that addresses all three. MAGIC is a four-stage pipeline that turns a single natural-language prompt into a runnable multi-scene game project: it plans a shared transition-aware intermediate representation, specifies each scene while enforcing portal reachability with a flood-fill validator, generates the scenes together with their transition scripts, and combines them into one project. Because existing single-scene fidelity metrics never execute a transition, we further introduce a transition-focused evaluation agent that runs each transition in play. On a new benchmark of 100 multi-scene cases, MAGIC produces an executable project for every case and reaches 0.99 precision, 0.95 recall, and 0.96 F1 on end-to-end transition identification; stage by stage, it recovers more ground-truth portals and yields markedly more navigable layouts than an LLM baseline and Holodeck. Our code is available at https://github.com/sereneee1201/MAGIC/.

Automated UML generation from sketches and images is gaining renewed attention with the rise of large language models and multimodal AI. However, reproducible evaluation remains difficult due to the lack of public datasets with executable groundtruth models. We present CAS2UML, a public dataset of 557 handdrawn UML diagrams, including 271 class diagrams and 286 activity diagrams, each paired with manually validated PlantUML code. We also provide a PlantUML-based validation tool and reusable scripts for checking the syntactic correctness and renderability of generated UML artifacts, enabling reproducible benchmarking of sketch-to-UML approaches. The dataset, validation tool, processing scripts, documentation, and demonstration video are publicly available at: Dataset: https://huggingface.co/datasets/Seym0n /cas2uml_hand-drawn_to_plantuml_dataset; Tool and Scripts: https://github.com/Seym0n/handwritten-uml-dataset; Video: https://www.youtube.com/watch?v=KQrYeGgT3hs.

Coding agents now produce a growing share of a team's code, while the reasoning behind each change -- the alternatives weighed, the constraints discovered, the approaches rejected -- is trapped in assistant transcripts that vanish with the session. Memory for this setting, the agentic development lifecycle (ADLC), is usually posed as one retrieval problem and built as machinery: tiered stores, memory graphs, compiled wikis, model-judged admission. We argue memory should instead be git-bound -- built into the repository's version control, inheriting the guarantees the machinery struggles to construct: ground truth from commits, freshness from rebuild, verification from the merge, containment from review. On this ledger we solve two problems separately, then combine them. Seed supply is closed as an eight-corpus retrieval study under a pre-registered ship discipline: five imported ranking mechanisms rejected, two kept, and a best configuration of ~0.31 pooled MRR -- ~60x the raw-transcript grep floor, ~15x an honest parsed-turn floor. Answer assembly is where ranking stops helping: single-shot retrieval scores only 0.07-0.20 answer-sufficiency on real developer questions, and ungated episode injection measurably degrades good answers. A router dispatches breadth to a git-anchored structural map, pointed lookups to confidence-gated episodes, and rationale to decision synthesis, which reconstructs why-arcs no single session contains (0.83 sufficiency on a young ~50k-LOC production system). Routed, the system answers at 382-980 tokens per question -- three orders of magnitude below the recorded history. Because ground truth is mined from commit-session links rather than annotated, every result is replicable on any user's own history at zero labeling cost. The remaining constraint is capture. Code, benchmark, and paper source: github.com/rekal-dev/rekal-cli.

Training terminal agents requires scalable executable supervision, yet synthesizing high-quality terminal tasks remains challenging. Each task couples an instruction, an initialized environment, a reference solution, and an executable verifier; if these artifacts are generated from inconsistent assumptions, the resulting task may be unsolvable or incorrectly evaluated. Meanwhile, multi-stage synthesis can discard the goals, dependencies, state transitions, and procedural constraints encoded in the original sources. We present FACET (Fine-grained Agentic Construction of Executable Tasks), a framework that addresses both information preservation and cross-artifact consistency. FACET reconstructs related agent skills into coherent, information-rich scenarios, then realizes and repairs the execution environment before generating the final task artifacts. The resulting container state serves as shared grounding for the instruction, solution, and verifier, while execution-based validation and targeted repair correct artifact-specific failures without unnecessarily regenerating valid components. FACET produces complex terminal tasks with dense executable checks, and successful trajectories collected from these tasks provide effective, data-efficient supervision. Fine-tuning models across multiple scales consistently improves performance on Terminal-Bench 2.1, while analyses of alternative generation schemes support the importance of environment-grounded construction for task validity and solution-verifier alignment. These results establish source-intent preservation and shared executable-state grounding as key principles for scalable terminal-task synthesis.

AI coding agents such as Claude Code and Gemini CLI increasingly extend themselves with third-party skills: markdown packages bundling natural-language instructions, executable scripts, and tool permissions. Because a skill is at once code and agent-facing instruction, it introduces a supply chain dependency whose risk is neither pure code nor pure prompt. Detection tools have never been measured against verified ground truth spanning this hybrid space, leaving their effectiveness unknown and wild-only evaluations biased. We present MalSkillBench, the first runtime-verified benchmark of malicious agent skills: 3,944 malicious skills labeled along a three-dimensional taxonomy of 108 cells. Of these, 3,214 come from a closed-loop Generate-Verify-Feedback pipeline admitting only samples whose malicious behavior fires inside a Docker sandbox under system-call monitoring and an LLM judge; we add 703 in-the-wild and 4,000 matched benign skills. Our measurements are consistent: code injection reaches 94.5% verification yield but prompt injection only 75.8%, the same fragility that later makes it hard to detect; the wild sample is narrow, dominated by one cryptocurrency-theft campaign (86.6% one behavior, 81% from two accounts) with a small but architecturally new tail attacking the agent control plane; the strongest skill-specific detector reaches 98.4% recall on code injection yet collapses on prompt-injection and agent-control attacks, and wild-only scoring swings the ranking by up to 66 recall points; supply-chain scanners and prompt-injection defenses each see only half of a skill, and no combination recovers the code-instruction relationship. Detecting malicious skills therefore requires reasoning jointly over task intent, code, and instructions. We release the dataset, pipeline, baselines, and results.

Land-use and land cover (LULC) analysis is critical in remote sensing, with wide-ranging applications across diverse fields such as agriculture, utilities, and urban planning. However, automating LULC map generation using machine learning is rendered challenging due to noisy labels. Typically, the ground truths (e.g. ESRI LULC, MapBioMass) have noisy labels that hamper the model's ability to learn to accurately classify the pixels. Further, these erroneous labels can significantly distort the performance metrics of a model, leading to misleading evaluations. Traditionally, the ambiguous labels are rectified using unsupervised algorithms. These algorithms struggle not only with scalability but also with generalization across different geographies. To overcome these challenges, we propose a zero-shot approach using the foundation model, Segment Anything Model (SAM), to automatically delineate different land parcels/regions and leverage them to relabel the unsure pixels by using the local label statistics within each detected region. We achieve a significant reduction in label noise and an improvement in the performance of the downstream segmentation model by approx 5% when trained with denoised labels.

AI pentesting agents are increasingly credible as offensive security systems, but current benchmarks still provide limited guidance on which will perform best in real-world targets. Existing evaluation protocols assess and optimize for predefined goals such as capture-the-flag, remote code execution, exploit reproduction, or trajectory similarity, in simplified or narrow settings. These tools are valuable for measuring bounded capabilities, yet they do not adequately capture the complexity, open-ended exploration, and strategic decision-making required in realistic pentesting. In this paper, we present a practical evaluation protocol that shifts assessment from task completion to validated vulnerability discovery, allowing evaluation in sufficiently complex targets spanning multiple attack surfaces and vulnerability classes. The protocol combines structured ground-truth with LLM-based semantic matching to identify vulnerabilities, bipartite resolution to score findings under realistic ambiguity, continuous ground-truth maintenance, repeated and cumulative evaluation of stochastic agents, efficiency metrics, and reduced-suite selection for sustainable experimentation. This protocol extends the state of the art by enabling a more realistic, operationally informative comparison of AI pentesting agents. To enable reproducibility, we also release expert-annotated ground truth and code for the proposed evaluation protocol: https://github.com/ethiack/ethibench.

AI agents deployed as persistent assistants must maintain correct beliefs as their information environment evolves. In practice, evidence is scattered across heterogeneous sources that often contradict one another, new information can invalidate earlier conclusions, and user preferences surface through corrections rather than explicit instructions. Existing benchmarks largely assume static, single-authority settings and do not evaluate whether agents can keep up with this complexity. We introduce ClawArena, a benchmark for evaluating AI agents in evolving information environments. Each scenario maintains a complete hidden ground truth while exposing the agent only to noisy, partial, and sometimes contradictory traces across multi-channel sessions, workspace files, and staged updates. Evaluation is organized around three coupled challenges: multi-source conflict reasoning, dynamic belief revision, and implicit personalization, whose interactions yield a 14-category question taxonomy. Two question formats, multi-choice (set-selection) and shell-based executable checks, test both reasoning and workspace grounding. The current release contains 64 scenarios across 8 professional domains, totaling 1{,}879 evaluation rounds and 365 dynamic updates. Experiments on five agent frameworks and five language models show that both model capability (15.4% range) and framework design (9.2%) substantially affect performance, that self-evolving skill frameworks can partially close model-capability gaps, and that belief revision difficulty is determined by update design strategy rather than the mere presence of updates. Code is available at https://github.com/aiming-lab/ClawArena.

Reference-free faithfulness metrics verify each atomic claim a model makes against ground truth, and are increasingly used to evaluate grounded generation. We show they share a blind spot: they measure only precision -- are the stated claims supported? -- and therefore reward abstention, since a model can score near-perfect faithfulness by saying almost nothing. We make this measurable using Formula 1 telemetry, a domain where strategic ground truth is derived deterministically and, crucially, completely: for each decision we know the full set of facts that mattered. This completeness -- absent in open-domain faithfulness benchmarks -- lets us measure recall (coverage of the relevant facts) exactly, alongside precision. On a multilingual (EN/ES/PT) benchmark of 7,253 decision instances spanning 150 races, the most precise frontier model covers under half of the relevant facts and ranks last by F1, so requiring coverage reorders the systems; the same effect reappears in a second complete-oracle domain (NOAA weather forecasts). A prompt ablation shows the low coverage is not an under-prompting artifact: explicitly asking models to be thorough does not close the gap. We pair faithfulness with coverage into a single score, validate the metric (controlled perturbation; agreement across a model-free regex extractor and a cross-family LLM extractor, system-level Spearman 1.0), and give a verifier-guided generation method that improves precision and recall without references. We release the benchmark, structured annotations, metric, baselines, and an interactive demo.

While LLMs exhibit remarkable fluency, their utility is often compromised by factual hallucinations and a lack of traceable provenance. Existing resources for grounding mitigate this but typically enforce a dichotomy: they offer either structured knowledge without textual context (e.g., knowledge bases) or grounded text with limited scale and linguistic coverage. To bridge this gap, we introduce FactNet, a massive, open-source resource designed to unify 1.7 billion atomic assertions with 3.01 billion auditable evidence pointers derived exclusively from 316 Wikipedia editions. Unlike recent synthetic approaches, FactNet employs a strictly deterministic construction pipeline, ensuring that every evidence unit is recoverable with byte-level precision. Extensive auditing confirms a high grounding precision of 92.1%, even in long-tail languages. Furthermore, we establish FactNet-Bench, a comprehensive evaluation suite for Knowledge Graph Completion, Question Answering, and Fact Checking. FactNet provides the community with a foundational, reproducible resource for training and evaluating trustworthy, verifiable multilingual systems.

Code LLMs still struggle with code execution reasoning, especially in smaller models. Existing methods rely on supervised fine-tuning (SFT) with teacher-generated explanations, primarily in two forms: (1) input-output (I/O) prediction chains and (2) natural-language descriptions of execution traces. However, intermediate execution steps cannot be explicitly verified during SFT, so the training objective can reduce to merely matching teacher explanations. Moreover, training data is typically collected without explicit control over task difficulty. We introduce ExecVerify, which goes beyond text imitation by incorporating verifiable white-box rewards derived from execution traces, including next-statement prediction and variable value/type prediction. Our work first builds a dataset with multiple difficulty levels via constraint-based program synthesis. Then, we apply reinforcement learning (RL) to reward correct answers about both intermediate execution steps and final outputs, aligning the training objective with semantic correctness at each execution step. Finally, we adopt a two-stage training pipeline that first enhances execution reasoning and then transfers to code generation. Experiments demonstrate that a 7B model trained with ExecVerify achieves performance comparable to 32B models on code reasoning benchmarks and improves pass@1 by up to 5.9\% on code generation tasks over strong post-training baselines.

While Large Language Models (LLMs) have substantially advanced text-to-code synthesis, many real programming tasks specify intent through visual artifacts such as screenshots, charts, vector drawings, videos, and interactive states. These tasks require models to connect visual perception to executable programs, because correctness depends not only on syntax but also on layout, data semantics, interaction behavior, and domain-specific constraints that apply after execution. This survey examines Multimodal Code Intelligence, covering systems that generate, edit, refine, or reason with code under visually grounded inputs and outputs. We first formulate the field by the role that code plays in each task, distinguishing code as a rendered artifact, an editable symbolic structure, a scientific representation, an intermediate reasoning trace, or an executable policy or tool interface. We then organize benchmarks and methods into four domains: Graphical User Interface, Scientific Visualization, Structured Graphics, and Frontier Tasks and Frameworks. This taxonomy connects mature artifact-generation problems to emerging agentic and unified settings and allows us to compare how different tasks treat evidence of correctness. Looking ahead, we argue that future research may benefit from four verification-centered directions. Multi-signal validation can combine complementary evidence of correctness, multi-state verification can test behavior across execution trajectories, cross-task transfer testing can probe reusable visual-code skills, and verifiable agent traces can reveal whether agent actions are grounded in visual evidence. Together, these directions may move this field from single-output imitation toward evidence-grounded executable systems. An ongoing project and resources are available on https://github.com/xjywhu/Awesome-Multimodal-LLM-for-Code{GitHub}.

With the emergence of large-scale vision language models (VLM), it is now possible to produce realistic-looking radiology reports for chest X-ray images. However, their clinical translation has been hampered by the factual errors and hallucinations in the produced descriptions during inference. In this paper, we present a novel phrase-grounded fact-checking model (FC model) that detects errors in findings and their indicated locations in automatically generated chest radiology reports. Specifically, we simulate the errors in reports through a large synthetic dataset derived by perturbing findings and their locations in ground truth reports to form real and fake findings-location pairs with images. A new multi-label cross-modal contrastive regression network is then trained on this dataset. We present results demonstrating the robustness of our method in terms of accuracy of finding veracity prediction and localization on multiple X-ray datasets. We also show its effectiveness for error detection in reports of SOTA report generators on multiple datasets achieving a concordance correlation coefficient of 0.997 with ground truth-based verification, thus pointing to its utility during clinical inference in radiology workflows.

Agent Skills package reusable instructions and assets for tool-using language-model agents. Progressive loading creates failure boundaries poorly represented by session-, model-, or tool-centric traces: a Skill can be discovered but not activated, activated without instructions, or appear successful without an independently verified outcome. We present Skill Runtime Intelligence, a passive runtime-intelligence system that reconstructs supported Skill-lifecycle stages across heterogeneous harnesses while preserving unsupported stages as unknown. Its Run Panorama separates immutable events, deterministic relations, inferred diagnoses, and controlled outcomes with four evidence grades; optional trace import and OTLP/HTTP export support existing observability deployments. Across six frozen repository profiles, three coding agents, and seven clean or fault-injected conditions, all 126 executions preserve source worktrees and each correlates to exactly one source session. Yet adapters expose three distinct semantics: no Skill runs; complete runs but no failure-like events; or failure-like events in every operational-failure and clean session. In a seven-template diagnostic study, semantic aliases and Panorama localize the same six non-clean boundaries but differ in exact/status behavior; both Raw views emit a failure status on all 18 clean cases, while Panorama emits none. A known-rule graph conforms to 126/126 frozen contracts, whereas a second model completes only 228/378 calls. These observations motivate executable adapter qualification and show that event presence is not boundary fidelity, composite exact scores mask distinct errors, and model explanations must not overwrite deterministic facts.

Vibe coding produces correct, executable code at speed, but leaves no record of the structural commitments, dependencies, or evidence behind it. Reviewers cannot determine what invariants were assumed, what changed, or why a regression occurred. This is not a generation failure but a control failure: the dominant artifact of AI-assisted development (code plus chat history) performs dimension collapse, flattening complex system topology into low-dimensional text and making systems opaque and fragile under change. We propose Agentic Consensus: a paradigm in which the consensus layer C, an operable world model represented as a typed property graph, replaces code as the primary artifact of engineering. Executable artifacts are derived from C and kept in correspondence via synchronization operators Phi (realize) and Psi (rehydrate). Evidence links directly to structural claims in C, making every commitment auditable and under-specification explicit as measurable consensus entropy rather than a silent guess. Evaluation must move beyond code correctness toward alignment fidelity, consensus entropy, and intervention distance. We propose benchmark task families designed to measure whether consensus-based workflows reduce human intervention compared to chat-driven baselines.

Enterprise workflows increasingly rely on agents for schema-guided extraction: given a document and a user-defined schema, the agent faithfully follows the schema to produce the correct output with source evidence as grounding metadata. We present ExtractBench, a benchmark for schema-guided extraction and, to our knowledge, the first to score value accuracy, record completeness at scale, grounding, and measured cost together. The evaluation system contains 4,869 pages across 370 enterprise documents, 8 business domains, and 67 document types, with clear tags differentiating their challenge scenarios. The scalable schema and ground-truth curation pipeline combines independent-system agreement for real documents, known values for synthetic lists, and human verification for forms. We report order-insensitive value F1 for value accuracy, plus two grounding metrics for source traceability: word- and page-level F1. Commercial VLMs perform well on short documents but often truncate record lists on long ones, while coding agents retain higher accuracy at much higher cost. LlamaExtract Agentic Plus ranks first on all three metrics, with accuracy comparable to coding agents at a fraction of the cost. Dataset and evaluation code are available on https://huggingface.co/datasets/llamaindex/ExtractBench{HuggingFace} and https://github.com/run-llama/ExtractBench{GitHub}.

Execution Accuracy (EX), the widely used metric for evaluating the effectiveness of Natural Language to SQL (NL2SQL) solutions, is becoming increasingly unreliable. It is sensitive to syntactic variation, ignores that questions may admit multiple interpretations, and is easily misled by erroneous ground-truth SQL. To address this, we introduce ROSE, an intent-centered metric that focuses on whether the predicted SQL answers the question, rather than consistency with the ground-truth SQL under the reference-dependent paradigm. ROSE employs an adversarial Prover-Refuter cascade: SQL Prover assesses the semantic correctness of a predicted SQL against the user's intent independently, while Adversarial Refuter uses the ground-truth SQL as evidence to challenge and refine this judgment. On our expert-aligned validation set ROSE-VEC, ROSE achieves the best agreement with human experts, outperforming the next-best metric by nearly 24% in Cohen's Kappa. We also conduct a largescale re-evaluation of 19 NL2SQL methods, revealing four valuable insights. We release ROSE and ROSE-VEC to facilitate more reliable NL2SQL research.

In the multi-commit development model, programmers complete tasks (e.g., implementing a feature) by organizing their work in several commits and packaging them into a commit-set. Analyzing data from developers using this model can be useful to tackle challenging developers' needs, such as knowing which features introduce a bug as well as assessing the risk of integrating certain features in a release. However, to do so one first needs to identify fix-inducing commit-sets. For such an identification, the SZZ algorithm is the most natural candidate, but its performance has not been evaluated in the multi-commit context yet. In this study, we conduct an in-depth investigation on the reliability and performance of SZZ in the multi-commit model. To obtain a reliable ground truth, we consider an already existing SZZ dataset and adapt it to the multi-commit context. Moreover, we devise a second dataset that is more extensive and directly created by developers as well as Quality Assurance (QA) engineers of Mozilla. Based on these datasets, we (1) test the performance of B-SZZ and its non-language-specific SZZ variations in the context of the multi-commit model, (2) investigate the reasons behind their specific behavior, and (3) analyze the impact of non-relevant commits in a commit-set and automatically detect them before using SZZ.

Large language models generate plausible code but cannot verify correctness. Existing multi-agent systems simulate execution or leave verification optional. We introduce execution-grounded verification as a first-class principle: every code change must survive sandboxed execution before propagation. We instantiate this principle in AGENTFORGE, a multi-agent framework where Planner, Coder, Tester, Debugger, and Critic agents coordinate through shared memory and a mandatory Docker sandbox. We formalize software engineering with LLMs as an iterative decision process over repository states, where execution feedback provides a stronger supervision signal than next-token likelihood. AGENTFORGE achieves 40.0\% resolution on SWE-BENCH Lite, outperforming single-agent baselines by 26--28 points. Ablations confirm that execution feedback and role decomposition each independently drive performance. The framework is open-source at https://github.com/raja21068/AutoCodeAI.

User instructions are often underspecified because humans rely on implicit assumptions about the surrounding environment. For large language model (LLM) agents operating in information-rich digital and physical environments, these assumptions cannot be inferred from the instruction alone; they must be recovered from the current state of tools, data, interfaces, and observations. Effective execution therefore requires agents to identify missing context, ground it in observed evidence, and carry it forward into subsequent actions. We show that current agents often fail to do so. They act from assumed rather than observed specifics, overlook information they could have gathered, and fail to incorporate evidence that has already been returned. Building on this insight, we propose ACCORD (Action-Conditioned Contextual Grounding), a simple and effective agent framework for adaptive grounding. Before each action, ACCORD actively probes the environment for missing information and integrates relevant context from the agent's trajectory that would otherwise be overlooked. Requiring no additional training or task-success signals, ACCORD improves task-goal completion on AppWorld by up to +20.6 points with GPT-5-mini, from 42.0% to 62.6%, compared to strong baselines. These gains persist with a substantially stronger base model (+10.8 with Claude-4.5-sonnet), an open-weight model (+10.1 with Qwen3.5-27B-FP8), and on the embodied AlfWorld benchmark (+7.4 success rate with GPT-5-mini).

The main issue with most evaluation schemes today is their "static" nature: the same problems are reused repeatedly, allowing for memorization, format exploitation, and eventual saturation. To measure genuine AI progress, we need evaluation that is robust by construction, not by post-hoc detection. In response, we propose VeRA (Verified Reasoning Data Augmentation), a framework that converts benchmark problems into executable specifications, comprising (i) a natural language template with placeholder slots, (ii) a coherent generator that samples valid configurations, and (iii) a deterministic verifier that validates parameters and calculates the corresponding correct answers for each configuration. From a single seed problem, VeRA automatically creates unlimited verified variants with reliable labels at near-zero marginal cost without human involvement. VeRA operates in two complementary modes. VeRA-E (equivalent) rewrites problems while keeping the underlying logic intact, useful for detecting memorization versus genuine reasoning. VeRA-H (hardened) systematically increases complexity while remaining verifiable, enabling reliable creation and labelling of fresh difficult tasks at the boundary of intelligence. Evaluating 16 frontier models with VeRA, we find: (i) VeRA-E improves evaluation quality and reveals contamination patterns. (ii) VeRA-H enables human-free generation of hard tasks with reliable labels. (iii) VeRA establishes verified benchmarks as a general paradigm. VeRA reconceptualizes benchmarks from static objects used until exhausted, to executable specifications generating fresh, verified instances on demand, enhancing robustness and cost-effectiveness for evaluation. With VeRA, we envision that evaluation in any verifiable domain can scale indefinitely without sacrificing label integrity. To stimulate future research, we have open-sourced all code and datasets.

Tool calling has emerged as a critical capability for AI agents to interact with the real world and solve complex tasks. While the Model Context Protocol (MCP) provides a powerful standardized framework for tool integration, there is a significant gap in benchmarking how well AI agents can effectively solve multi-step tasks using diverse MCP tools in realistic, dynamic scenarios. In this work, we present LiveMCP-101, a benchmark of 101 carefully curated real-world queries, refined through iterative LLM rewriting and manual review, that require coordinated use of multiple MCP tools including web search, file operations, mathematical reasoning, and data analysis. Moreover, we introduce a novel evaluation approach that leverages ground-truth execution plans rather than raw API outputs, better reflecting the evolving nature of real-world environments. Experiments show that even frontier LLMs achieve a success rate below 60\%, highlighting major challenges in tool orchestration. Detailed ablations and error analysis further reveal distinct failure modes and inefficiencies in token usage, pointing to concrete directions for advancing current models. LiveMCP-101 sets a rigorous standard for evaluating real-world agent capabilities, advancing toward autonomous AI systems that reliably execute complex tasks through tool use.

Multimodal Large Language Models (MLLMs) have significantly advanced document understanding, yet current Doc-VQA evaluations score only the final answer and leave the supporting evidence unchecked. This answer-only approach masks a critical failure mode: a model can land on the correct answer while grounding it in the wrong passage -- a critical risk in high-stakes domains like law, finance, and medicine, where every conclusion must be traceable to a specific source region. To address this, we introduce CiteVQA, a benchmark that requires models to return element-level bounding-box citations alongside each answer, evaluating both jointly. CiteVQA comprises 1,897 questions across 711 PDFs spanning seven domains and two languages, averaging 40.6 pages per document. To ensure fidelity and scalability, the ground-truth citations are generated by an automated pipeline-which identifies crucial evidence via masking ablation-and are subsequently validated through expert review. At the core of our evaluation is Strict Attributed Accuracy (SAA), which credits a prediction only when the answer and the cited region are both correct. Auditing 20 MLLMs reveals a pervasive Attribution Hallucination: models frequently produce the right answer while citing the wrong region. The strongest system (Gemini-3.1-Pro-Preview) achieves an SAA of only 76.0, and the strongest open-source MLLM reaches just 22.5. Ultimately, towards trustworthy document intelligence, CiteVQA exposes a reliability gap that answer-only evaluations overlook, providing the instrumentation needed to close it. Our repository is available at https://github.com/opendatalab/CiteVQA.

In real home deployments, household agents must often operate from a complete household scene and a situated household request, rather than from a clean task specification. Such requests require agents to identify task-relevant entities, recover intended task conditions, and resolve ordering constraints from the surrounding scene context. We formalize this capability as full-scene household reasoning: given a complete household scene and a situated household request, an agent must infer executable task structure before producing a grounded skill-level action sequence. This setting is challenging because complete household scenes contain substantial task-irrelevant information, making direct complete-scene prompting inefficient and error-prone. In practical deployment, this challenge is further amplified by privacy and local compute constraints, which favor compact open-weight models with limited long-context reasoning ability. We propose TaskGround, a training-free and model-agnostic Ground-Infer-Execute framework that grounds complete scenes into compact task-relevant scene slices, infers executable task structure, and compiles it into grounded skill-level action sequences. To evaluate this setting, we introduce FullHome, a human-validated evaluation suite of 400 household tasks spanning diverse home-scale environments and both goal-oriented and process-constrained requirements. On FullHome, TaskGround improves task success rates by large margins across both proprietary and open-weight models. Notably, it makes Qwen3.5-9B competitive with GPT-5 under direct complete-scene prompting while reducing total input-token cost by up to 18x. Our results identify executable task-structure inference as a central bottleneck in full-scene household reasoning and show that structured grounding can make compact local models substantially more effective for practical household deployment.

Vision Language Models (VLMs) are typically evaluated with Visual Question Answering (VQA) tasks which assess a model's understanding of scenes. Good VQA performance is taken as evidence that the model will perform well on a broader range of tasks that require both visual and language inputs. However, scene-aware VQA does not fully capture input biases or assess hallucinations caused by a misalignment between modalities. To address this, we propose a Cross-modal Alignment Similarity Test (CAST) to probe VLMs for self-consistency across modalities. This test involves asking the models to identify similarities between two scenes through text-only, image-only, or both and then assess the truthfulness of the similarities they generate. Since there is no ground-truth to compare against, this evaluation does not focus on objective accuracy but rather on whether VLMs are internally consistent in their outputs. We argue that while not all self-consistent models are capable or accurate, all capable VLMs must be self-consistent.

Natural-language autoencoders score explanations of hidden activations by reconstruction: an explanation is deemed faithful if the activation can be regenerated from it. The test is structurally insensitive to individual false claims: if flipping a claim does not change the reconstruction, the claim is never penalized. We show the test is passed in two ways, neither faithful. On a released Qwen-2.5-7B verbalizer, explanations reconstruct well above chance while ~2% of specific claims are reconstruction-dependent, so the score tracks gist, not specific facts. Under exact synthetic ground truth, the standard recipe develops co-adapted private codes (false wording the reconstruction depends on) in 5/5 runs, and fixes that leave the target model unchanged do not help. We contribute two audit protocols, the grounded-vs-true cross and the evaluator swap, and RECAP (Readable Encodings via Co-trained Auxiliary Predictors): linear heads trained alongside the target model to keep designated content decodable. On RECAP-trained sandbox models, fresh verbalizers state the designated content truly and the codes vanish, at a +0.001-nat cost. This replicates on a pretrained Pythia-160M: the content becomes reliably probe-decodable, though a fresh verbalizer conveys it only in part (truth 0.44-0.46 vs a near-zero control). For interpretability, high reconstruction does not certify individual claims. For AI safety, RECAP makes designated internal content independently checkable against probes rather than asserted by prose a model can game: an independent probe scores the verbalizer's true claims above its false ones (AUC 0.96, vs 0.82 without RECAP). Against an adversary that edits an explanation to maximize the reconstruction score while lying (suppressing ~87% of its lie penalty), the RECAP probe still flags the lies (AUC 0.95) while the control probe collapses to chance (0.51).

Data cleaning is a time-consuming and error-prone manual process, even with modern workflow tools such as OpenRefine. We present AutoDCWorkflow, an LLM-based pipeline for automatically generating data-cleaning workflows. The pipeline takes a raw table and a data analysis purpose, and generates a sequence of OpenRefine operations designed to produce a minimal, clean table sufficient to address the purpose. Six operations correspond to common data quality issues, including format inconsistencies, type errors, and duplicates. To evaluate AutoDCWorkflow, we create a benchmark with metrics assessing answers, data, and workflow quality for 142 purposes using 96 tables across six topics. The evaluation covers three key dimensions: (1) Purpose Answer: can the cleaned table produce a correct answer? (2) Column (Value): how closely does it match the ground truth table? (3) Workflow (Operations): to what extent does the generated workflow resemble the human-curated ground truth? Experiments show that Llama 3.1, Mistral, and Gemma 2 significantly enhance data quality, outperforming the baseline across all metrics. Gemma 2-27B consistently generates high-quality tables and answers, while Gemma 2-9B excels in producing workflows that closely resemble human-annotated versions.

Graphical User Interface (GUI) grounding aims to translate natural language instructions into executable screen coordinates, enabling automated GUI interaction. Nevertheless, incorrect grounding can result in costly, hard-to-reverse actions (e.g., erroneous payment approvals), raising concerns about model reliability. In this paper, we introduce SafeGround, an uncertainty-aware framework for GUI grounding models that enables risk-aware predictions through calibrations before testing. SafeGround leverages a distribution-aware uncertainty quantification method to capture the spatial dispersion of stochastic samples from outputs of any given model. Then, through the calibration process, SafeGround derives a test-time decision threshold with statistically guaranteed false discovery rate (FDR) control. We apply SafeGround on multiple GUI grounding models for the challenging ScreenSpot-Pro benchmark. Experimental results show that our uncertainty measure consistently outperforms existing baselines in distinguishing correct from incorrect predictions, while the calibrated threshold reliably enables rigorous risk control and potentials of substantial system-level accuracy improvements. Across multiple GUI grounding models, SafeGround improves system-level accuracy by up to 5.38% percentage points over Gemini-only inference.

Recent graphical user interface (GUI) grounders have significantly advanced single-shot accuracy on standard benchmarks, yet their performance degrades sharply on small targets, densely packed controls and out-of-distribution interfaces. We attribute this gap to a paradigmatic limitation shared by existing approaches: none of them treats a produced coordinate as a hypothesis to be reflected upon and revised under new visual evidence. This manifests as three coupled issues: 1) Lack of post-hoc reflection. The prediction is frozen at the moment of emission, leaving no internal mechanism to challenge or refine it. 2) Visual evidence decoupled from the prediction. The auxiliary visual evidence is gathered to support the upcoming coordinate rather than to scrutinise the one already committed to. 3) Refinement over views, not over predictions. The iterative zoom-in refines the inspected region instead of inheriting a previous coordinate as a spatial prior to be corrected. In this paper, we propose LookAgain, a closed-loop GUI grounder driven by post-prediction visual reflection. LookAgain reformulates grounding as a multi-turn predict-look-again-refine process with two primitives: "locate" posts a coordinate hypothesis, renders a marker on the image and appends a local patch of the predicted region. It anchors the next reasoning step to the previous prediction as a spatial prior; "confirm" accepts or reject the hypothesis and terminates the procedure. We train the LookAgain grounder with SFT on constructed reflective trajectories as a cold start, followed by GRPO with terminal grounding correctness as the sole reward. Extensive experiments show that LookAgain consistently improves performance on both refusal-aware and general GUI grounding benchmarks, achieving state-of-the-art results. Comprehensive ablations further verify the effectiveness of the proposed framework.

We introduce FAITHSCORE (Faithfulness to Atomic Image Facts Score), a reference-free and fine-grained evaluation metric that measures the faithfulness of the generated free-form answers from large vision-language models (LVLMs). The FAITHSCORE evaluation first identifies sub-sentences containing descriptive statements that need to be verified, then extracts a comprehensive list of atomic facts from these sub-sentences, and finally conducts consistency verification between fine-grained atomic facts and the input image. Meta-evaluation demonstrates that our metric highly correlates with human judgments of faithfulness. We collect two benchmark datasets (i.e. LLaVA-1k and MSCOCO-Cap) for evaluating LVLMs instruction-following hallucinations. We measure hallucinations in state-of-the-art LVLMs with FAITHSCORE on the datasets. Results reveal that current systems are prone to generate hallucinated content unfaithful to the image, which leaves room for future improvements. Further, we find that current LVLMs despite doing well on color and counting, still struggle with long answers, relations, and multiple objects.

The remarkable performance of Multimodal Large Language Models (MLLMs) has unequivocally demonstrated their proficient understanding capabilities in handling a wide array of visual tasks. Nevertheless, the opaque nature of their black-box reasoning processes persists as an enigma, rendering them uninterpretable and struggling with hallucination. Their ability to execute intricate compositional reasoning tasks is also constrained, culminating in a stagnation of learning progression for these models. In this work, we introduce Fact, a novel paradigm designed to generate multimodal rationales that are faithful, concise, and transferable for teaching MLLMs. This paradigm utilizes verifiable visual programming to generate executable code guaranteeing faithfulness and precision. Subsequently, through a series of operations including pruning, merging, and bridging, the rationale enhances its conciseness. Furthermore, we filter rationales that can be transferred to end-to-end paradigms from programming paradigms to guarantee transferability. Empirical evidence from experiments demonstrates the superiority of our method across models of varying parameter sizes, significantly enhancing their compositional reasoning and generalization ability. Our approach also reduces hallucinations owing to its high correlation between images and text.

We present ExCyTIn-Bench, the first benchmark to Evaluate an LLM agent x on the task of Cyber Threat Investigation through security questions derived from investigation graphs. Real-world security analysts must sift through a large number of heterogeneous alert signals and security logs, follow multi-hop chains of evidence, and compile an incident report. With the developments of LLMs, building LLM-based agents for automatic thread investigation is a promising direction. To assist the development and evaluation of LLM agents, we construct a dataset from a controlled Azure tenant that covers 8 simulated real-world multi-step attacks, 57 log tables from Microsoft Sentinel and related services, and 589 automatically generated questions. We leverage security logs extracted with expert-crafted detection logic to build threat investigation graphs, and then generate questions with LLMs using paired nodes on the graph, taking the start node as background context and the end node as answer. Anchoring each question to these explicit nodes and edges not only provides automatic, explainable ground truth answers but also makes the pipeline reusable and readily extensible to new logs. This also enables the automatic generation of procedural tasks with verifiable rewards, which can be naturally extended to training agents via reinforcement learning. Our comprehensive experiments with different models confirm the difficulty of the task: with the base setting, the average reward across all evaluated models is 0.249, and the best achieved is 0.368, leaving substantial headroom for future research. Code and data are coming soon!

GUI grounding, which maps natural-language instructions to actionable UI elements, is a core capability of GUI agents. Prior works largely treats instructions as a static proxy for user intent, overlooking the impact of instruction diversity and quality on grounding performance. Through a careful investigation of existing grounding datasets, we find a 23.3% flaw rate in their instructions and show that inference-time exploitation of instruction diversity yields up to a substantial 76% relative performance improvement. In this paper, we introduce the Instruction-as-Reasoning paradigm, treating instructions as dynamic analytical pathways that offer distinct perspectives and enabling the model to select the most effective pathway during reasoning. To achieve this, we propose a two-stage training framework: supervised fine-tuning (SFT) on synthesized, diverse instructions to instill multi-perspective reasoning, followed by reinforcement learning (RL) to optimize pathway selection and composition. Our resulting models, UI-Ins-7B and UI-Ins-32B, achieve state-of-the-art results on five challenging grounding benchmarks and exhibit emergent reasoning, selectively composing and synthesizing novel instruction pathways at inference. In particular, UI-Ins-32B attains the best grounding accuracy, scoring 87.3% on UI-I2E-Bench, 57.0% on ScreenSpot-Pro, and 84.9% on MMBench-GUI L2. Furthermore, our model demonstrates strong agentic potential, achieving a 74.1% success rate on AndroidWorld using UI-Ins-7B as the executor. Our in-depth analysis reveals additional insights such as how reasoning can be formulated to enhance rather than hinder grounding performance, and how our method mitigates policy collapse in the SFT+RL framework. All code and model checkpoints will be publicly released in https://github.com/alibaba/UI-Ins.

Language models of code (LMs) work well when the surrounding code in the vicinity of generation provides sufficient context. This is not true when it becomes necessary to use types or functionality defined in another module or library, especially those not seen during training. LMs suffer from limited awareness of such global context and end up hallucinating, e.g., using types defined in other files incorrectly. Recent work tries to overcome this issue by retrieving global information to augment the local context. However, this bloats the prompt or requires architecture modifications and additional training. Integrated development environments (IDEs) assist developers by bringing the global context at their fingertips using static analysis. We extend this assistance, enjoyed by developers, to the LMs. We propose a notion of monitors that use static analysis in the background to guide the decoding. Unlike a priori retrieval, static analysis is invoked iteratively during the entire decoding process, providing the most relevant suggestions on demand. We demonstrate the usefulness of our proposal by monitoring for type-consistent use of identifiers whenever an LM generates code for object dereference. To evaluate our approach, we curate PragmaticCode, a dataset of open-source projects with their development environments. On models of varying parameter scale, we show that monitor-guided decoding consistently improves the ability of an LM to not only generate identifiers that match the ground truth but also improves compilation rates and agreement with ground truth. We find that LMs with fewer parameters, when guided with our monitor, can outperform larger LMs. With monitor-guided decoding, SantaCoder-1.1B achieves better compilation rate and next-identifier match than the much larger text-davinci-003 model. The datasets and code will be released at https://aka.ms/monitors4codegen .

Fact verification is essential for ensuring the reliability of LLM applications. In this study, we evaluate 12 pre-trained LLMs and one specialized fact-verifier, including frontier LLMs and open-weight reasoning LLMs, using a collection of examples from 14 fact-checking benchmarks. We share three findings intended to guide future development of more robust fact verifiers. First, we highlight the importance of addressing annotation errors and ambiguity in datasets, demonstrating that approximately 16\% of ambiguous or incorrectly labeled data substantially influences model rankings. Neglecting this issue may result in misleading conclusions during comparative evaluations, and we suggest using a systematic pipeline utilizing LLM-as-a-judge to help identify these issues at scale. Second, we discover that frontier LLMs with few-shot in-context examples, often overlooked in previous works, achieve top-tier performance. We therefore recommend future studies include comparisons with these simple yet highly effective baselines. Lastly, despite their effectiveness, frontier LLMs incur substantial costs, motivating the development of small, fine-tuned fact verifiers. We show that these small models still have room for improvement, particularly on instances that require complex reasoning. Encouragingly, we demonstrate that augmenting training with synthetic multi-hop reasoning data significantly enhances their capabilities in such instances. We release our code, model, and dataset at https://github.com/just1nseo/verifying-the-verifiers

LLM-based dialogue assistants have become mainstream tools for software developers, yet current evaluation benchmarks focus exclusively on functional correctness. This leaves a critical gap in assessing the quality and accuracy of these conversations when handling Non-Functional Requirements (NFRs), which are inherently vague, context-dependent, and involve many parts of a program. Evaluating how well these systems support collaborative reasoning about NFRs requires methods that go beyond single-turn accuracy to capture both the correctness of the system's outputs and the quality of the multi-turn interaction. In this paper, we investigate the accuracy and quality of multi-turn conversations between developers and an LLM-based agent in the domain of Health Insurance Portability and Accountability Act (HIPAA) regulatory compliance. We hired 49 programmers to interact with GitHub Copilot to assess 148 HIPAA-derived NFRs against the iTrust codebase, a system designed to comply with HIPAA regulations, across three dimensions: requirement satisfaction level, reasoning, and code localization. We find that developers tend to agree with LLM assessments, but accuracy against expert ground truth is low. We model user satisfaction and find that longer system responses and more information-providing turns negatively affect user satisfaction, whereas proactive interactions positively affect it. Our findings provide insights for designing LLM-based dialogue systems that support NFR assessment.

Recently, given the docstring for the target problem and the target function signature, large language models (LLMs) have been used not only to generate source code, but also to generate test cases, consisting of test inputs and assertions (e.g., in the form of checking an actual output against the expected output). However, as shown by our empirical study on assertions generated by four LLMs for the HumanEval benchmark, over 62% of the generated assertions are incorrect (i.e., failed on the ground-truth problem solution). To detect incorrect assertions (given the docstring and the target function signature along with a sample of example inputs and outputs), in this paper, we propose a new approach named DeCon to effectively detect incorrect assertions via LLM-generated postconditions for the target problem (a postcondition is a predicate that must always be true just after the execution of the ground-truth problem solution). Our approach requires a small set of I/O examples (i.e., a sample of example inputs and outputs) for the target problem (e.g., the I/O examples included in the docstring for a target problem in HumanEval). We use the given I/O examples to filter out those LLM-generated postconditions that are violated by at least one given I/O example. We then use the remaining postconditions to detect incorrect assertions as those assertions that violate at least one remaining postcondition. Experimental results show that DeCon can detect averagely more than 64% (63% and 65.5% detected by GPT-3.5 and GPT-4, respectively) incorrect assertions generated by four state-of-the-art LLMs, and DeCon can also improve the effectiveness of these LLMs in code generation by 4% in terms of Pass@1. In addition, although DeCon might filter out correct assertions, the fault-finding ability of the remaining correct assertions decreases only slightly.

Embodied intelligence, a grand challenge in artificial intelligence, is fundamentally constrained by the limited spatial understanding and reasoning capabilities of current models. Prevailing efforts to address this through enhancing Vision-Language Models (VLMs) are trapped in a dilemma: template-based datasets are scalable but structurally rigid, while manual annotation is linguistically diverse but unscalable and, critically, computationally imprecise. We introduce SPRITE, a novel framework that overcomes this dilemma by leveraging simulators and large models to programmatically synthesize scalable, diverse, and high-quality spatial reasoning data. The core innovation of SPRITE is to reframe ground-truth generation as a code-generation task. We utilize LLMs to compile complex spatial questions into executable programs, which are then verified against high-precision scene meta-information extracted from simulators. This ensures our ground truth is both computationally precise and verifiable, while the generative power of LLMs provides vast linguistic diversity. Leveraging this pipeline, we have curated a dataset encompassing 3 simulators, 11k+ scenes, and 300k+ image/video instruction-tuning pairs. We demonstrate that a VLM trained on our data achieves significant performance gains on multiple spatial benchmarks and outperforms other open-source datasets of equivalent size. Furthermore, a scalability analysis confirms our hypothesis that overcoming the low-diversity nature of traditional template methods is essential for building robust, generalizable spatial intelligence. We will make the SPRITE framework code and the full 300k+ dataset publicly available to facilitate future research in spatial intelligence.

State-of-the-art fact-checking systems combat misinformation at scale by employing autonomous LLM-based agents to decompose complex claims into smaller sub-claims, verify each sub-claim individually, and aggregate the partial results to produce verdicts with justifications (explanatory rationales for the verdicts). The security of these systems is crucial, as compromised fact-checkers, which tend to be easily underexplored, can amplify misinformation. This work introduces Fact2Fiction, the first poisoning attack framework targeting such agentic fact-checking systems. Fact2Fiction mirrors the decomposition strategy and exploits system-generated justifications to craft tailored malicious evidences that compromise sub-claim verification. Extensive experiments demonstrate that Fact2Fiction achieves 8.9\%--21.2\% higher attack success rates than state-of-the-art attacks across various poisoning budgets. Fact2Fiction exposes security weaknesses in current fact-checking systems and highlights the need for defensive countermeasures.

Vision-language models (VLMs) excel on visual question answering and multimodal reasoning benchmarks. Yet their capability on ultra-resolution images - where critical evidence is tiny, subtle, spatially distant, or distributed - remains unclear. Existing evaluations largely report final-answer accuracy, offering limited insight into whether models acquire and integrate the necessary visual evidence. We introduce UltraVR, a diagnostic benchmark for evidence-grounded visual reasoning over ultra-resolution images. UltraVR spans four high-value scenarios: CCTV surveillance, remote sensing (RS), whole-slide image (WSI) pathology, and industrial anomaly detection (AD). These domains pose complementary challenges: fine-grained object grounding in crowded CCTV scenes, long-range spatial comparison in RS, multi-scale evidence navigation in WSI, and subtle irregularity detection in repetitive industrial layouts. Beyond standard QA triples, each instance includes a structured ground-truth chain of thought with step-level questions, intermediate answers, and reasoning labels. These labels decompose reasoning into evidence grounding, local perception, quantification, evidence integration, and decision inference, enabling process-level diagnosis over black-box scoring. Using UltraVR, we evaluate frontier VLMs and show that current models remain far from reliable on ultra-resolution reasoning. Importantly, the structured annotations allow us to localize failures across the visual-to-decision pipeline: errors concentrate in evidence grounding and local perception, while downstream inference often recovers when intermediate visual facts are supplied. These findings demonstrate UltraVR as a diagnostic testbed for measuring not only whether VLMs answer correctly, but where their ultra-resolution reasoning process breaks.

Process reward models (PRMs) that provide dense, step-level feedback have shown promise for reinforcement learning, yet their adoption remains limited by the need for expensive step-level annotations or ground truth references. We propose SPARK: a three-stage framework where in the first stage a generator model produces diverse solutions and a verifier model evaluates them using parallel scaling (self-consistency) and sequential scaling (meta-critique). In the second stage, we use these verification outputs as synthetic training data to fine-tune generative process reward models, which subsequently serve as reward signals during training. We show that aggregating multiple independent verifications at the step level produces training data for process reward models that surpass ground-truth outcome supervision, achieving 67.5 F1 on ProcessBench (a benchmark for identifying erroneous steps in mathematical reasoning) compared to 66.4 for reference-guided training and 61.9 for GPT-4o. In the final stage, we apply our generative PRM with chain-of-thought verification (PRM-CoT) as the reward model in RL experiments on mathematical reasoning, and introduce format constraints to prevent reward hacking. Using Qwen2.5-Math-7B, we achieve 47.4% average accuracy across six mathematical reasoning benchmarks, outperforming ground-truth-based RLVR (43.9%). Our work enables reference-free RL training that exceeds ground-truth methods, opening new possibilities for domains lacking verifiable answers or accessible ground truth.

Existing benchmarks for LLM coding agents primarily evaluate final outcomes. While useful for measuring overall capability, these metrics provide limited visibility and often miss defects that arise during execution. We present ProcCtrlBench, a benchmark for execution-process evaluation in LLM coding agents. ProcCtrlBench organizes recurrent execution defects into a reusable ontology covering 11 defect types in 4 categories, and evaluates agent trajectories through standardized process evidence rather than final outcomes alone. To support comparison across heterogeneous agents, ProcCtrlBench standardizes raw logs into a unified trajectory representation and reports calibrated scorecards over process-level findings. In addition, ProcCtrlBench uses control preservation as a way to quantify execution-process quality, capturing whether execution remains interpretable, interruptible, correctable, reversible, and able to hand back authority when needed. We evaluate ProcCtrlBench on 200 cases sampled from three benchmarks: AndroidBench, TerminalBench, and SWE-bench-Verified. Results show that ProcCtrlBench can be instantiated with useful reliability, provides more stable semantics than direct thresholding, and reveals meaningful differences in execution quality that are often overlooked by conventional outcome-based evaluation.

Executable software engineering data is valuable for training SWE agents, but scaling it remains difficult for two reasons: only a small fraction of real repository changes yield verifiable, high-signal task instances, and naively building repository-specific environments quickly becomes the dominant systems cost. We present SWE-Next, an execution-grounded framework for scalable SWE task and trajectory collection. On the data side, SWE-Next mines real merged pull requests, executes candidate base/merged commit pairs, and retains only those that produce strict test improvements without regressions, yielding self-verifying instances. It also applies strict submission gating so that collected trajectories remain evidence-driven rather than speculative. On the systems side, SWE-Next introduces reusable repo-quarter profiles, which reuse the same environment across nearby commits in time while keeping each task run separate and reproducible. Using only 30 hours and 639GB of environment storage, SWE-Next processes 3,971 seed repositories and 102,582 candidate commit pairs mined from real merged PRs to construct a dataset of 2,308 self-verifying instances. Experiments show that SWE-Next improves downstream pass@1 with fewer or comparable training trajectories, indicating that its gains come not from a stronger trajectory generator, but from higher-signal execution-grounded supervision and more efficient data collection.

Social deduction games have become a popular testbed for probing reasoning, deception, coordination, and belief modeling in Large Language Model (LLM) agents. However, most environments are scored only by game outcomes such as win rates and largely remain to text-only interaction, making it difficult to tell whether an agent's language is actually grounded in what it perceived and did, or to identify the failure modes underlying its behavior. To address this gap, we introduce QUACK, an open-source environment and evaluation framework for auditing the grounding of agent language in multimodal social reasoning. QUACK evaluates agents at three levels: game outcomes, behavioral trajectories, and utterance-level consistency. Its core Statement Verification Pipeline reconstructs each agent's ground-truth trajectory from engine logs and checks every discussion claim against it, automatically flagging spatial hallucination, unsupported accusation, deception collapse, and language-action inconsistency. Evaluating three frontier VLMs in both homogeneous and cross-model adversarial settings, we find that even the strongest agent hallucinates 15.1% of its verifiable spatial claims and makes over half of its accusations without grounded evidence. We release the full engine, evaluation framework, toolkit, and logs at https://github.com/AAAAA-Academia-Attractions/QUACK.

Agents powered by large language models (LLMs) are increasingly adopted in the software industry, contributing code as collaborators or even autonomous developers. As their presence grows, it becomes important to assess the current boundaries of their coding abilities. Existing agentic coding benchmarks, however, cover a limited task scope, e.g., bug fixing within a single pull request (PR), and often rely on non-executable evaluations or lack an automated approach for continually updating the evaluation coverage. To address such issues, we propose FeatureBench, a benchmark designed to evaluate agentic coding performance in end-to-end, feature-oriented software development. FeatureBench incorporates an execution-based evaluation protocol and a scalable test-driven method that automatically derives tasks from code repositories with minimal human effort. By tracing from unit tests along a dependency graph, our approach can identify feature-level coding tasks spanning multiple commits and PRs scattered across the development timeline, while ensuring the proper functioning of other features after the separation. Using this framework, we curated 200 challenging evaluation tasks and 3825 executable environments from 24 open-source repositories in the first version of our benchmark. Empirical evaluation reveals that the state-of-the-art agentic model, such as Claude 4.5 Opus, which achieves a 74.4% resolved rate on SWE-bench, succeeds on only 11.0% of tasks, opening new opportunities for advancing agentic coding. Moreover, benefiting from our automated task collection toolkit, FeatureBench can be easily scaled and updated over time to mitigate data leakage. The inherent verifiability of constructed environments also makes our method potentially valuable for agent training.

Large Language Models (LLMs) are fluent but prone to hallucinations, producing answers that appear plausible yet are unsupported by available evidence. This failure is especially problematic in high-stakes domains where decisions must be justified by verifiable information. We introduce EvidenceRL, a reinforcement learning framework that enforces evidence adherence during training. EvidenceRL scores candidate responses for grounding (entailment with retrieved evidence and context) and correctness (agreement with reference answers) and optimizes the generator using Group Relative Policy Optimization (GRPO). We evaluate across two high-stakes domains, cardiac diagnosis and legal reasoning, where EvidenceRL consistently improves evidence grounding and faithfulness without sacrificing task accuracy. On cardiac diagnosis, F1@3 increases from 37.0 to 54.5 on Llama-3.2-3B while grounding (G_{max}@3) rises from 47.6 to 78.2; hallucinations drop nearly 5times and evidence-supported diagnoses increase from 31.8\% to 61.6\%. On legal reasoning, EvidenceRL raises Faithfulness from 32.8\% to 67.6\% on Llama-3.1-8B, demonstrating consistent behavioral change across domains. Our code is open-sourced at https://github.com/Wizaaard/EvidenceRL.git.

Verifying the authenticity of satellite imagery has become increasingly critical given advances in generative artificial intelligence. Highly realistic synthetic imagery produced for malicious purposes (deepfakes) can have major consequences in the remote sensing domain, where this data is a fundamental source of information for science applications, planning, logistics, and monitoring. The remote sensing community lacks high-quality, fine-grained manipulation datasets suitable for training and evaluating detection and image forensics algorithms. Existing datasets are lacking and those that do exist either provide no ground truth masks for evaluating manipulation localization, or consist of entire images generated by GANs or diffusion models, which are inadequate for measuring localization performance. To address this gap, we describe a preliminary dataset construction process and prototype benchmark dataset for satellite image manipulation detection and localization. The dataset contains 60 images total, with 30 images carefully manipulated using three manipulation types including copy-paste splicing and diffusion model inpainting, and 30 authentic images. Each image is accompanied by a ground-truth mask and acquisition metadata, enabling both pixel-level localization metrics, image metadata studies, and analyses of how manipulation detection performance relates to image collection parameters. We describe the dataset construction process and present this initial release to support further research in image forensics and geospatial deepfake detection. The prototype dataset can be downloaded at https://huggingface.co/datasets/geodf/fmow-fake-small.

Despite the rapid progress of Multimodal Large Language Models (MLLMs), their ability to perform reliable visual grounding in high-stakes clinical software environments remains underexplored. Existing GUI benchmarks largely focus on isolated, single-step grounding queries, overlooking the sequential, workflow-driven reasoning required in real-world medical interfaces, where tasks evolve across independent steps and dynamic interface states. We introduce MedSPOT, a workflow-aware sequential grounding benchmark for clinical GUI environments. Unlike prior benchmarks that treat grounding as a standalone prediction task, MedSPOT models procedural interaction as a sequence of structured spatial decisions. The benchmark comprises 216 task-driven videos with 597 annotated keyframes, in which each task consists of 2 to 3 interdependent grounding steps within realistic medical workflows. This design captures interface hierarchies, contextual dependencies, and fine-grained spatial precision under evolving conditions. To evaluate procedural robustness, we propose a strict sequential evaluation protocol that terminates task assessment upon the first incorrect grounding prediction, explicitly measuring error propagation in multi-step workflows. We further introduce a comprehensive failure taxonomy, including edge bias, small-target errors, no prediction, near miss, far miss, and toolbar confusion, to enable systematic diagnosis of model behavior in clinical GUI settings. By shifting evaluation from isolated grounding to workflow-aware sequential reasoning, MedSPOT establishes a realistic and safety-critical benchmark for assessing multimodal models in medical software environments. Code and data are available at: https://github.com/Tajamul21/MedSPOT.

Decompilation aims to restore compiled code to human-readable source code, but struggles with details like names and structure. Large language models (LLMs) show promise for programming tasks, motivating their application to decompilation. However, there does not exist any open-source LLM for decompilation. Moreover, existing decompilation evaluation systems mainly consider token-level accuracy and largely ignore code executability, which is the most important feature of any program. Therefore, we release the first open-access decompilation LLMs ranging from 1B to 33B pre-trained on 4 billion tokens of C source code and the corresponding assembly code. The open-source LLMs can serve as baselines for further development in the field. To ensure practical program evaluation, we introduce Decompile-Eval, the first dataset that considers re-compilability and re-executability for decompilation. The benchmark emphasizes the importance of evaluating the decompilation model from the perspective of program semantics. Experiments indicate that our LLM4Decompile has demonstrated the capability to accurately decompile 21% of the assembly code, which achieves a 50% improvement over GPT-4. Our code, dataset, and models are released at https://github.com/albertan017/LLM4Decompile

Before letting an agent operate over real context, can you prove it used the right evidence? GroundEval turns that question into a deterministic test of what the agent searched, fetched, cited, and was permitted to access. In one case study, two frontier LLM judges scored a plausible agent response above 0.85. But the trace told a different story: the agent had never retrieved the artifact its answer depended on, yielding a GroundEval score of 0.000. We introduce GroundEval, a judge-free framework for evaluating agents against grounded, time-bounded, and access-controlled evidence. GroundEval uses a domain configuration to generate questions, lets the agent choose how to answer, and then scores both the final answer and the recorded trajectory that produced it. The benchmark targets three failures that LLM-as-judge evaluation struggles to detect: whether an agent checked before claiming absence, reasoned only from evidence available to the actor at the relevant time, and used the correct causal mechanism rather than a plausible one. These correspond to three tracks: Silence, Perspective, and Counterfactual. GroundEval exposes when plausible answers rest on invalid evidence paths, and produces structured per-question diagnostics that pair tool activity with the agent's turn-level narration, making each score inspectable rather than merely reported. What our case studies turned up is that this gap isn't some rare corner case. It's exactly the blind spot that final-answer and judge-based scoring were never built to catch.

Understanding and reasoning about code semantics is essential for enhancing code LLMs' abilities to solve real-world software engineering (SE) tasks. Although several code reasoning benchmarks exist, most rely on synthetic datasets or educational coding problems and focus on coarse-grained reasoning tasks such as input/output prediction, limiting their effectiveness in evaluating LLMs in practical SE contexts. To bridge this gap, we propose CodeSense, the first benchmark that makes available a spectrum of fine-grained code reasoning tasks concerned with the software engineering of real-world code. We collected Python, C and Java software projects from real-world repositories. We executed tests from these repositories, collected their execution traces, and constructed a ground truth dataset for fine-grained semantic reasoning tasks. We then performed comprehensive evaluations on state-of-the-art LLMs. Our results show a clear performance gap for the models to handle fine-grained reasoning tasks. Although prompting techniques such as chain-of-thought and in-context learning helped, the lack of code semantics in LLMs fundamentally limit models' capabilities of code reasoning. Besides dataset, benchmark and evaluation, our work produced an execution tracing framework and tool set that make it easy to collect ground truth for fine-grained SE reasoning tasks, offering a strong basis for future benchmark construction and model post training. Our code and data are located at https://codesense-bench.github.io/.

Post-training via Supervised Fine-Tuning (SFT) and Reinforcement Learning (RL) is crucial for enhancing reasoning in Multimodal Large Language Models (MLLMs), yet existing paradigms often reach a performance bottleneck due to the limitations of static data. While current methods leverage self-reflection or self-evolution to push these boundaries, they still suffer from cognitive drift and hallucinated reasoning paths caused by low-quality synthetic data. To address these challenges, we propose Anchor Evolution (AnE), a new paradigm that integrates truth-anchored data curation and model evolution, achieving faithful and steady performance gains at the reasoning frontier. Specifically, we propose Truth Anchor Expansion, which pinpoints the model failing frontier via trajectory rollouts and leverages ground-truth databases to retrieve high-fidelity anchors for faithful data curation. Subsequently, we introduce the Scaffold-Stripping Mechanism to internalize reasoning capabilities. This mechanism first anchors reasoning paths via scaffold-augmented supervision to mitigate the learning complexity and distribution drift of direct SFT on raw data, then leverages RL to strip the scaffold template, thereby effectively transitioning the reasoning paths into intrinsic model capabilities. Experimental results on multimodal reasoning benchmarks show that our method substantially advances the model performance frontier, improving the base model by 10.3\% across eight multimodal benchmarks and achieving state-of-the-art results. The code will be made publicly available.

Safety alignment in open-weight language models is trivially removable: abliteration projects a refusal-mediating direction out of the weights in minutes, and no release-time defense we are aware of prevents it durably. What cannot be prevented can be deceived. Our defense, decoy hardening ("Fool's Gold"), concedes the refusal strip and poisons its payoff: once refusal is stripped, most answers to hazardous operational requests are confident, fluent decoys whose critical elements are falsified. Decoys are trained inside a differentiable simulation of the attack, expressing only in the attacked state; a refusal pin and benign leash hold clean-state behavior to the original. We instantiate it on seven models from five families (9B-122B, dense and mixture-of-experts). On the six models passing our pre-registered efficacy gate, 0.51-0.90 of attacked-state responses to held-out prompts are decoys, +0.27-0.84 attributable to the defense; all six stay within registered benign-behavior and capability budgets; the seventh (smaller) fails the gate (boundary case). Rates replicate on a frozen test split or untouched strata. The claim is epistemic: without independent ground truth, no observation surface we tested separates falsified answers from correct ones - on external red-team benchmarks' CBRNE-adjacent slice, the defended 122B is fatally wrong on 0.82-0.86 of matched-quality answers vs at most 0.10 undefended. Repeated sampling does not restore trust: element-wise consensus at K=64 reconstructs a fully usable procedure on 0.083-0.625 of prompts where the instrument validates, vs 0.58-0.96 undefended, with no label-free way to tell the regimes apart; on the weakest such model the claim is per-draw only. We evaluate chemical and biological hazards; the defense does not address in-context jailbreaks and protects only the initially released defended weights.

The rapid progress of frontier large language models has led to widespread benchmark saturation, limiting the ability of existing datasets to differentiate model capabilities or provide useful training signal. For instance, on LiveCodeBench, frontier models achieve over 99% Pass@1 on easy splits and exceed 90% Pass@1 on average across difficulty levels. Constructing new, challenging datasets typically requires substantial human effort, creating a bottleneck for progress. We introduce BenchEvolver, a solution-centric evolutionary framework that automatically transforms existing coding problems into harder variants. Rather than generating problems from scratch, BenchEvolver evolves reference solutions through structured transformations and derives corresponding statements and tests from the evolved solutions. This design grounds generation in executable semantics, enabling scalable construction of high-quality, diverse, and difficult tasks with verifiable correctness. Applying BenchEvolver to LiveCodeBench and SciCode, we obtain evolved tasks that are substantially harder while maintaining validity, reference correctness, and diversity. We further curate LiveCodeBench-Plus, a 91-problem benchmark combining evolved and difficult original LCB-v6 tasks, where frontier-model Pass@1 ranges from 27.5% to 62.6%, restoring clear discrimination among strong coding models. Importantly, evolved tasks remain challenging even for the model that generates them, enabling self-improvement. We further show that RL on evolved LCB tasks improves held-out coding performance: for gpt-oss-20b, seed+evolved training achieves +8.7 and +8.3 Pass@1 gains on LCB v6 Hard and LCB-Pro Easy, exceeding seed-only gains by 70.7% and 34.8%, respectively. Our results show that BenchEvolver can convert saturated benchmarks into frontier-level evaluation suites and reusable training signal.

Formal verification is the next frontier for ensuring the correctness of code generated by Large Language Models (LLMs). While methods that co-generate code and formal specifications in formal languages, like Dafny, can, in principle, prove alignment with user intent, progress is bottlenecked by specification quality evaluation. Current benchmarks rely on matching against ground-truth specifications, a manual and expertise-intensive process that has limited existing datasets to a few hundred simple problems and also suffers from a reliability issue. To address this, we introduce VeriEquivBench, a new benchmark with 2,389 complex algorithmic problems that probe the limitations of current models in both code generation and formal reasoning. Our evaluation framework replaces ground-truth matching with a formally grounded metric, the equivalence score, and rigorously verifies the quality of generated specifications and code. Our results show that generating formally verifiable code remains a profound challenge for state-of-the-art LLMs. This underscores both the difficulty of the task and the need for benchmarks like VeriEquivBench to drive progress toward scalable and reliable coding agents.

Self-improving LLM agents increasingly learn from experience without updating any weights. Each episode is stored in an external memory, scored, and retrieved for similar future tasks to shape later behavior. Viewed through a reward lens, the stored score is a proxy reward for an implicit, non-parametric policy. Each retrieved episode then becomes a policy-improvement step whose reliability hinges on how that score is produced. In deployment, ground-truth labels are unavailable, so the stored reward is at best an LLM assessment. This substitution creates a failure mode, the *Echo Gap*, across the memory-based self-improving agents and model families studied. Incorrect episodes receive inflated rewards; thus, the agent preferentially reuses the very mistakes it has most confident in. Because the error compounds through memory rather than averaging out and the confirming judge's errors remain correlated with the original self-grading bias, so it cannot identify which memories are overvalued. The missing property is formalized as the *Error-Independence Assumption* (EIA), which we prove is a *necessary* condition for correcting the inflation, not merely a description of a good verifier: a usable signal must track truth *and* decorrelate its error from the memory bias, and the recoverable payoff is a closed-form function of exactly those two quantities. We further show the inflation compounds not only when retrieval ranks by the stored score but also under plain similarity retrieval which is the regime the deployed agent uses. Finally, the answer-free de-inflation algorithm LUCID delivers a consistent end-to-end gain on the BIRD text-to-SQL benchmark. It raises execution accuracy to 56.9%, above both a Memento-style self-graded agent (54.0%, a +2.9-point mean gain across seeds) and a memory-less agent of identical architecture (52.4%).

LLM agents choose tools and arguments from context that mixes user requests, tool outputs, retrieved records, memory, and untrusted text. Evidence can be relevant without being authorized to determine a decision, so a correct action need not be grounded only in permitted evidence. We introduce a target-specific authorization audit that labels context factors separately for each tool and argument target. Its primary test holds the task, proposition, position, and policy fixed while changing only the proposition's source authority. We then test behavior when valid evidence is weakened and use context-subset interactions as a secondary localization diagnostic. Across 450 controlled next-action tasks and multiple open-weight LLM families, trusted and untrusted variants produce different actions in 5.4 percent of competing cases versus 1.7 percent of supporting cases. Under controlled degradation, unauthorized competition is retained in a full-correct, mixed-error, clean-correct pattern in 2.4 percent of comparisons, with a 95 percent confidence interval from 2.1 to 3.0 percent. These are controlled stress-set rates, not deployment prevalence. The models respond to textual source-authority cues, but this does not prevent untrusted evidence from influencing their actions.

Evidence-grounded reasoning requires more than attaching retrieved text to a prediction: a model should make decisions that depend on whether the provided evidence supports the target claim. In practice, this often fails because supervision is weak, evidence is only loosely tied to the claim, and evaluation does not test evidence dependence directly. We introduce case-grounded evidence verification, a general framework in which a model receives a local case context, external evidence, and a structured claim, and must decide whether the evidence supports the claim for that case. Our key contribution is a supervision construction procedure that generates explicit support examples together with semantically controlled non-support examples, including counterfactual wrong-state and topic-related negatives, without manual evidence annotation. We instantiate the framework in radiology and train a standard verifier on the resulting support task. The learned verifier substantially outperforms both case-only and evidence-only baselines, remains strong under correct evidence, and collapses when evidence is removed or swapped, indicating genuine evidence dependence. This behavior transfers across unseen evidence articles and an external case distribution, though performance degrades under evidence-source shift and remains sensitive to backbone choice. Overall, the results suggest that a major bottleneck in evidence grounding is not only model capacity, but the lack of supervision that encodes the causal role of evidence.

Publishing a large language model (LLM) benchmark on the Internet risks contaminating future LLMs: the benchmark may be unintentionally (or intentionally) used to train or select a model. A common mitigation is to keep the benchmark private and let participants submit their models or predictions to the organizers. However, this strategy will require trust in a single organization and still permits test-set overfitting through repeated queries. To overcome this issue, we propose a way to publish benchmarks without completely disclosing the ground-truth answers to the questions, while still maintaining the ability to openly evaluate LLMs. The main underlying idea is to reduces the best possible accuracy, i.e., Bayes accuracy, by injecting randomness to the answers by preparing several logically correct answers, and only include one of them as the solution in the benchmark. Not only is this helpful to keep us from disclosing the ground truth, but this also offers a test for detecting data contamination. In principle, even fully capable models should not surpass the Bayes accuracy. If a model surpasses this ceiling despite this expectation, this is a strong signal of data contamination. We present experimental evidence that our method can detect data contamination accurately on a wide range of benchmarks, models, and training methodologies.

Trustworthy deployment of LLM-based agents in software systems requires evaluating how they perform on application-specific workflows, with enough granularity to localize where they succeed and fail. Yet existing agent evaluation mechanisms are limited: benchmarks have low construct validity for application-specific workflows and environments, and replica evaluation environments are expensive and prone to drift. We propose Copy-on-Write (CoW) Scoring, a framework that evaluates agent operations directly within application environments using a PostgreSQL-level Copy-on-Write mechanism to isolate agent writes. CoW Scoring produces session- and operation-level scores that highlight where agents' database write operations succeed and fail in a given application environment, enabling inexpensive evaluation and iteration on agent harnesses and tool surfaces. We demonstrate the framework on Plane, an open-source project-management platform, where analysis surfaced specific issues in the tool surface, and corresponding fixes produced measurable improvements on affected models. Python library: https://github.com/trail-ml/agent-cow-python

Large Language Models (LLMs) are one of the most promising developments in the field of artificial intelligence, and the software engineering community has readily noticed their potential role in the software development life-cycle. Developers routinely ask LLMs to generate code snippets, increasing productivity but also potentially introducing ownership, privacy, correctness, and security issues. Previous work highlighted how code generated by mainstream commercial LLMs is often not safe, containing vulnerabilities, bugs, and code smells. In this paper, we present a framework that leverages testing and static analysis to assess the quality, and guide the self-improvement, of code generated by general-purpose, open-source LLMs. First, we ask LLMs to generate C code to solve a number of programming tasks. Then we employ ground-truth tests to assess the (in)correctness of the generated code, and a static analysis tool to detect potential safety vulnerabilities. Next, we assess the models ability to evaluate the generated code, by asking them to detect errors and vulnerabilities. Finally, we test the models ability to fix the generated code, providing the reports produced during the static analysis and incorrectness evaluation phases as feedback. Our results show that models often produce incorrect code, and that the generated code can include safety issues. Moreover, they perform very poorly at detecting either issue. On the positive side, we observe a substantial ability to fix flawed code when provided with information about failed tests or potential vulnerabilities, indicating a promising avenue for improving the safety of LLM-based code generation tools.

Tool-augmented LLM agents raise new security risks: tool executions can introduce runtime-only behaviors, including prompt injection and unintended exposure of external inputs (e.g., environment secrets or local files). While existing scanners often focus on static artifacts, analyzing runtime behavior is challenging because directly executing untrusted tools can itself be dangerous. We present MCP-SandboxScan, a lightweight framework motivated by the Model Context Protocol (MCP) that safely executes untrusted tools inside a WebAssembly/WASI sandbox and produces auditable reports of external-to-sink exposures. Our prototype (i) extracts LLM-relevant sinks from runtime outputs (prompt/messages and structured tool-return fields), (ii) instantiates external-input candidates from environment values, mounted file contents, and output-surfaced HTTP fetch intents, and (iii) links sources to sinks via snippet-based substring matching. Case studies on three representative tools show that MCP-SandboxScan can surface provenance evidence when external inputs appear in prompt/messages or tool-return payloads, and can expose filesystem capability violations as runtime evidence. We further compare against a lightweight static string-signature baseline and use a micro-benchmark to characterize false negatives under transformations and false positives from short-token collisions.

Solving complex reasoning tasks is a key real-world application of agents. Thanks to the pretraining of Large Language Models (LLMs) on code data, recent approaches like CodeAct successfully use code as LLM agents' action, achieving good results. However, CodeAct greedily generates the next action's code block by relying on fragmented thoughts, resulting in inconsistency and instability. Moreover, CodeAct lacks action-related ground-truth (GT), making its supervision signals and termination conditions questionable in multi-turn interactions. To address these issues, we first introduce a simple yet effective end-to-end code generation paradigm, CodeProgram, which leverages code's systematic logic to align with global reasoning and enable cohesive problem-solving. Then, we propose Tree-of-Code (ToC), which self-grows CodeProgram nodes based on the executable nature of the code and enables self-supervision in a GT-free scenario. Experimental results on two datasets using ten popular zero-shot LLMs show ToC remarkably boosts accuracy by nearly 20% over CodeAct with less than 1/4 turns. Several LLMs even perform better on one-turn CodeProgram than on multi-turn CodeAct. To further investigate the trade-off between efficacy and efficiency, we test different ToC tree sizes and exploration mechanisms. We also highlight the potential of ToC's end-to-end data generation for supervised and reinforced fine-tuning.

Multi-hop QA benchmarks often reward Large Language Models (LLMs) for spurious correctness, where models reach correct answers through invalid intermediate reasoning. We propose SAFE, an LLM-as-verifier framework for evidence-grounded multi-hop QA. Rather than judging only the final answer after generation, SAFE verifies reasoning during generation by checking intermediate steps against the provided passages and previous reasoning trajectory. To make this process checkable, SAFE decomposes reasoning into atomic, evidence-grounded units represented with Knowledge Graph (KG) triples. At train-time, SAFE verifies benchmark supervision under KG-grounded constraints and constructs reliable verifier training data. At inference-time, an external verifier checks each generated step, identifies invalid reasoning, and provides correction feedback before errors propagate. Across three multi-hop QA benchmarks, SAFE improves accuracy by 8.8 pp on average. These results show that evidence-grounded multi-hop QA benefits from shifting LLM-based evaluation from post-hoc answer judgment to stepwise reasoning verification.

We propose claim-level falsification as a principle for test-time scaling and instantiate it through Claim-Level Reliability Assessment (CLR), a training-free framework that reallocates test-time compute from additional solution sampling to targeted verification. Since whole-trace evaluation often obscures decisive errors due to signal dilution from routine tokens, CLR condenses each reasoning trace into a compact set of decision-critical claims, thereby isolating its logical anchors. Furthermore, recognizing the inherent difficulty of generating entirely correct solutions under fixed model capabilities, CLR shifts the focus to semantic falsification. This approach exploits a fundamental asymmetry between solution construction and claim refutation. Constructing a valid solution requires a flawless reasoning path, whereas refuting an incorrect claim requires identifying only a single decisive flaw. This targeted search for negative evidence systematically compresses the survival space of high-confidence incorrect traces, effectively suppressing erroneous consensus via nonlinear reliability scoring. Across four LLMs and four reasoning benchmarks under matched budgets, CLR generally improves upon pass@1 and self-consistency. On GPT-OSS-20B/CMIMC25, for instance, CLR exceeds pass@1 by 27.15 percentage-points and raises self-consistency accuracy from 77.50\% to 82.19\% with 37.0\% fewer tokens.

LLM reasoning traces suffer from complex flaws -- *Step Internal Flaws* (logical errors, hallucinations, etc.) and *Step-wise Flaws* (overthinking, underthinking), which vary by sample. A natural approach would be to provide ground-truth labels to guide LLMs' reasoning. Contrary to intuition, we show that this yields no improvement in reasoning ability. We then propose CRAFT, a unified framework that mitigates both types of Step flaws, which builds a Reasoning Knowledge Graph (RKG) based on the consensus parts of multiple candidate traces, and synthesizes a high-quality trace through topological generation. Our approach improves label-prediction accuracy by 10+% on average, and consistently outperforms all baselines across both logical and mathematical reasoning benchmarks. Further, detailed benchmark evaluation proves that our method also improves the quality of LLMs' reasoning traces in multiple dimensions.

AI agents that execute tasks via tool calls frequently hallucinate results - fabricating tool executions, misstating output counts, or presenting inferences as facts. Recent approaches to verifiable AI inference rely on zero-knowledge proofs, which provide cryptographic guarantees but impose minutes of proving time per query, making them impractical for interactive agents. We propose NabaOS, a lightweight verification framework inspired by Indian epistemology (Nyaya Shastra), which classifies every claim in an LLM response by its epistemic source (pramana): direct tool output (pratyaksha), inference (anumana), external testimony (shabda), absence (abhava), or ungrounded opinion. Our runtime generates HMAC-signed tool execution receipts that the LLM cannot forge, then cross-references claims against these receipts to detect hallucinations in real time. We evaluate on NyayaVerifyBench, a new benchmark of 1,800 agent response scenarios across four languages with injected hallucinations of six types. NabaOS detects 94.2% of fabricated tool references, 87.6% of count misstatements, and 91.3% of false absence claims, with <15ms verification overhead per response. For deep delegation (agents performing multi-step web tasks), our cross-checking protocol catches 78.4% of URL fabrications via independent re-fetching. We compare against five approaches: zkLLM (cryptographic proofs, 180s/query), TOPLOC (locality-sensitive hashing), SPEX (sampling-based proof of execution), tensor commitments, and self-consistency checking. NabaOS achieves the best cost-latency-coverage trade-off for interactive agents: 94.2% coverage at <15ms versus zkLLM's near-perfect coverage at 180,000ms. For interactive agents, practical receipt-based verification provides better cost-benefit than cryptographic proofs, and epistemic classification gives users actionable trust signals rather than binary judgments.

While recent LLM-based terminal agents have demonstrated promising capabilities, the scarcity of high-quality, executable training data remains a critical bottleneck. Existing synthesis pipelines typically scale by retrofitting surface-level artifacts into tasks, frequently yielding ambiguous instructions, shallow execution paths, and brittle tests that provide weak learning signals. To overcome this, we introduce CLI-Universe, a principled synthesis engine that constructs terminal-agent tasks. CLI-Universe generates candidate tasks by sampling combinations across a multi-dimensional capability taxonomy (domain, skill type, capability, and engineering pillar), then grounds each candidate through evidence-guided deep research over real-world technical materials. To ensure rigorous supervision, validated blueprints are instantiated into Dockerized environments and subjected to a multi-stage executable verification pipeline featuring rubric-gated test construction, hint-conditional filtering, and strict fail-to-pass checking. Across the full pipeline, from candidate generation to verification, approximately two-thirds of candidates are discarded, retaining only those that are genuine, verifiable, and non-trivially challenging. To validate our framework, we instantiate a highly distilled dataset of 6,000 trajectories called CLI-Universe-6K. Remarkably, fine-tuning Qwen3-32B on CLI-Universe-6K achieves 33.4% on Terminal-Bench 2.0. This sets a new state-of-the-art for models trained on open-source data at or below 32B parameters, and outperforms several models an order of magnitude larger, demonstrating the profound data efficiency of structured, high-fidelity synthesis.

Code executability plays a vital role in software debugging and testing (e.g., detecting runtime exceptions or assertion violations). However, code execution, especially partial or arbitrary code execution, is a non-trivial task due to missing definitions and complex third-party dependencies. To make partial code (such as code snippets posted on the web or code fragments deep inside complex software projects) executable, the existing study has proposed a machine learning model to predict the undefined element types and inject the pre-defined dummy values into execution. However, the performance of their tool is limited due to its simply designed dummy values and the inability to continue learning. In this paper, we design and implement a novel framework, named SelfPiCo (Self Guided Partial Code Executor), to dynamically guide partial code execution by incorporating the open-source LLM (i.e., Code Llama) within an interactive loop. Particularly, SelfPiCo leverages few-shot in-context learning and chain-of-thought reasoning to elicit human knowledge and logical reasoning based on fine-tuning the Code Llama model. SelfPiCo continuously learns from code execution results and refines its predictions step after step. Our evaluations demonstrate that SelfPiCo can execute 72.7% and 83.3% of all lines in the open-source code and Stack Overflow snippets, outperforming the most recent state-of-the-art Lexecutor by 37.9% and 33.5%, respectively. Moreover, SelfPiCo successfully detected 18 and 33 runtime type error issues by executing the partial code from eight GitHub software projects and 43 Stack Overflow posts, demonstrating the practical usage and potential application of our framework in practice.

Agentic LLM coding tools compress long session histories into compaction summaries that subsequent sessions inherit as ground truth. This paper documents a failure mode in Claude Code where partial standard output from timed-out commands (exit code 143) is recorded in compaction summaries as confirmed results, propagating false positives across sessions and model versions without re-verification. The underlying mechanism is a conflation of observation and persistence, where information that appeared in the terminal is treated as equivalent to information written to durable storage. This finding extends the analysis of LLM self-evaluation failures reported in prior work on non-determinism in LLM-as-judge grading by showing that agentic tools exhibit analogous reliability deficits when reporting on their own operational outcomes. The failure has direct implications for any workflow that relies on agentic session continuity for data processing, scientific computation, or multi-step automation.

Agent benchmarks should reflect what users actually ask deployed agents to do, yet existing benchmarks often miss key realism properties of real developer-agent sessions. We introduce RealClawBench, a live benchmark framework built from real OpenClaw sessions to capture the distribution, diversity, and real-world difficulty of deployed agent use. Real user requests are challenging to benchmark because they often depend on local execution environments, involve implicit or underspecified intent, and require nontrivial verification. RealClawBench addresses these challenges with two core mechanisms: reconstructed execution environments and deterministic verifiable scorers, which together convert real sessions into reproducible, automatically scored tasks. The resulting release contains 281 executable tasks sampled from a much larger real-session pool while preserving the source distribution, with maximum final-vs-source Jensen-Shannon divergence of 0.0448. Evaluating 14 contemporary models shows that the best system solves only 65.8% of tasks, revealing substantial headroom on realistic developer-agent workloads. By turning real deployed sessions into controlled evaluation instances, RealClawBench provides a practical path toward benchmarks that better measure agent capability in actual use. Code is available at:https://anonymous.4open.science/r/real-claw-bench-582B.

Code Agents have achieved remarkable advances in recent years, exhibiting strong capabilities across a wide range of software engineering tasks. However, their misuse often produces bloated and disorganized code that impairing readability, extensibility, and robustness. Despite this risk, existing benchmarks largely evaluate functional correctness rather than long-term maintainability of code agents. In this paper, we propose SmellBench, an extensible code refactoring benchmark that proactively injects code smells into clean code snippets from real-world repositories. This design enables the generation of controlled, high-quality, and diverse refactoring cases with human-written ground truth. Specifically, it contains 294 cases spanning 7 popular smell types, 3 difficulty levels, 2 instruction settings across 7 real-world repositories. We further design 3 evaluation aspects covering functional correctness, localization ability, and refactoring quality assessment. Experiments with 2 popular agents and 6 large langauge models (LLMs) show that the best combination - Qwen Code + Claude Sonnet 4.5 - achieved only a 50.34 score of smell elimination. Further analysis reveals that this gap arises from a focus on local code smells and a lack of cross-file understanding, which hinders comprehensive smell elimination.

Agent skills extend LLM agents with reusable procedures, tools, and domain-specific workflows, but their safety depends on resolving dependencies among interacting instructions. We introduce SkillLogic, a framework for analyzing logical relations in skill files and constructing executable tests from them. Our taxonomy covers eight relation types, including preconditions that gate valid actions, constraints that limit how allowed actions may be performed, and fallbacks that specify recovery behavior after failure. Using SkillLogic, we scan over 5000 public skills and find that 70% contain at least one logical relation. We then construct SLBench, an 86-case executable benchmark from high-confidence, high-impact, and locally testable relations. Evaluating Codex and Claude Code across six LLM backbones shows unsafe rates up to 70%, with violations leading to privacy leaks, unsafe configuration changes, and incomplete cleanup. The human audit attributes failures to both agent capability gaps and low-salience skill text. We further show that SLGuard, a lightweight inference-time scaffold, reduces violations by 63% on targeted cases. Our results establish logical-relation following as a distinct reliability challenge for skill-guided agents.

The ability of CodeLLMs to generate executable and functionally correct code at the repository-level scale remains largely unexplored. We introduce RepoExec, a novel benchmark for evaluating code generation at the repository-level scale. RepoExec focuses on three main aspects: executability, functional correctness through automated test case generation with high coverage rate, and carefully crafted cross-file contexts to accurately generate code. Our work explores a controlled scenario where developers specify necessary code dependencies, challenging the model to integrate these accurately. Experiments show that while pretrained LLMs outperform instruction-tuned models in correctness, the latter excel in utilizing provided dependencies and demonstrating debugging capabilities. We also introduce a new instruction-tuned dataset that focuses on code dependencies and demonstrate that CodeLLMs fine-tuned on our dataset have a better capability to leverage these dependencies effectively. RepoExec aims to provide a comprehensive evaluation of code functionality and alignment with developer intent, paving the way for more reliable and applicable CodeLLMs in real-world scenarios. The dataset and source code can be found at~https://github.com/FSoft-AI4Code/RepoExec.

LLM-based code review agents promise scalable, always-on review, yet current systems suffer from two intertwined weaknesses: (1) non-determinism--unbounded tool use makes review outcomes unstable, and (2) context locality--the reviewer's access remains bounded to the diff, capping discoverable issue depth. Both give rise to three challenges: misaligned context retrieval, a coherence-efficiency trade-off in multi-file pull requests, and hallucinated comments that erode trust. To address these, we introduce OpenCodeReview, built on deterministic engineering for uncertain agents: rather than granting maximal freedom, we inject determinism at three deliberate pipeline points. Rule-Guided Dispatch uses a multi-layer rule system to deterministically select files and review criteria, eliminating variability of agent-driven triage. Grounded File Review replaces free-form exploration with a curated tool set exposed through a ReAct loop, while file-level parallel SubAgents balance context coherence against efficiency and recover cross-file dependencies on demand. Independent Reflection introduces a falsification-first filter under an asymmetric information boundary--the reflector sees only the diff, not the agent's tool-augmented exploration--removing hallucinated comments without self-reinforcing bias, improving precision while preserving recall. On AACR-Bench (200 real-world PRs, 10 languages, 1,505 expert-verified comments), OpenCodeReview outperforms mainstream coding agents (e.g., Claude Code and Codex) across six LLM backends, achieving up to 2.17x higher SEM-F1 (25.10% vs. 11.57%) while consuming 5-15x fewer tokens. We open-source OpenCodeReview at https://github.com/alibaba/open-code-review.

Autonomous coding agents increasingly execute multi-step software work, but lifecycle states such as reviewed, tested, DONE, and ready-to-merge remain claims unless supported by current evidence. We present Proof-or-Stop Lifecycle Control, a method that permits lifecycle transitions only when fresh, tracked-source-state-bound, mechanically verifiable evidence satisfies the relevant gate. The method treats agent outputs as claims rather than lifecycle state, and uses proof operationally to mean gate-admissible evidence under a stated trust model, not semantic program correctness. We evaluate an open-source implementation through mechanism tests, a powered control-policy ablation, and operated self-application evidence. The unattended-loop engine passed 10 of 10 scenarios with zero false-DONE, and local-key receipt bundles rejected 18 tamper classes with zero false accepts. In a 9,240-cell ablation, the pre-registered A4 versus A2-prime comparison reduced visible-pass/hidden-fail amplification from 31 of 1,800 injected cells under a compute-budgeted naive loop to 2 of 1,800 under the gated loop, a 1.6 percentage-point improvement in not-amplified rate with a 95 percent confidence interval of [0.8, 2.5]. A near-compute A3 versus A4 comparison, 14 of 1,800 versus 2 of 1,800, indicates that the gain is associated with enforcing review as a lifecycle gate rather than merely adding a reviewer. The self-application corpus contains 565 stories and 1,007 review findings, with 94.8 percent resolved, plus a 68-row high/critical cross-vendor exhibit. These results support Proof-or-Stop as a model-agnostic, host-neutral control layer for deciding which autonomous-agent claims a lifecycle may act on. The evaluation is limited to one model family, 24 ablation tasks, and a self-hosted corpus.

Deployed language models must produce outputs that are both correct and format-compliant. We study this structured-output reliability gap using two mathematical benchmarks -- GSM8K and MATH -- as a controlled testbed: ground truth is unambiguous and the output contract is strict (JSON with required fields). We evaluate three 7-9B models under five prompting strategies and report output accuracy -- the joint event of mathematical correctness and valid JSON structure -- as the primary metric. A systematic format failure emerges: NAIVE prompting (no system prompt) achieves up to 85% task accuracy on GSM8K but 0% output accuracy across all models and datasets. REFERENCE prompting (a minimal hand-written JSON format prompt) fares little better, yielding 0% output accuracy for two of four models tested. Constrained decoding enforces syntactic validity but incurs 3.6x-8.2x latency overhead and in several settings degrades task performance substantially. To overcome this limitation, we developed AloLab, an iterative system-prompt optimizer (meta-agent: Claude Sonnet 4.5) requiring only black-box API access to the target model; it reaches 84-87% output accuracy on GSM8K and 34-40% on MATH across five independent runs per model, with 29/30 paired McNemar comparisons against the best static prompt significant at p < 0.05, at near-NAIVE inference latency and without model fine-tuning. The same format failure extends to GPT-4o (OpenAI, 2024), a proprietary closed-source model: REFERENCE achieves 0% output accuracy due to systematic markdown-fence wrapping, while AloLab reaches 95.2% [94.8, 95.6]. An ablation replacing the Sonnet 4.5 meta-agent with Claude 3 Haiku reduces mean output accuracy to 61.0% and increases run-to-run standard deviation from <1 pp to 21.8 pp, confirming that meta-agent capability is a primary driver of optimization quality.

Visual reasoning is challenging, requiring both precise object grounding and understanding complex spatial relationships. Existing methods fall into two camps: language-only chain-of-thought approaches, which demand large-scale (image, query, answer) supervision, and program-synthesis approaches which use pre-trained models and avoid training, but suffer from flawed logic and erroneous grounding. We propose an annotation-free training framework that improves both reasoning and grounding. Our framework uses AI-powered verifiers: an LLM verifier refines LLM reasoning via reinforcement learning, while a VLM verifier strengthens visual grounding through automated hard-negative mining, eliminating the need for ground truth labels. This design combines the strengths of modern AI systems: advanced language-only reasoning models for decomposing spatial queries into simpler subtasks, and strong vision specialist models improved via performant VLM critics. We evaluate our approach across diverse spatial reasoning tasks, and show that our method improves visual reasoning and surpasses open-source and proprietary models, while with our improved visual grounding model we further outperform recent text-only visual reasoning methods. Project webpage: https://glab-caltech.github.io/valor/

Decompilation transforms compiled code back into a high-level programming language for analysis when source code is unavailable. Previous work has primarily focused on enhancing decompilation performance by increasing the scale of model parameters or training data for pre-training. Based on the characteristics of the decompilation task, we propose two methods: (1) Without fine-tuning, the Self-Constructed Context Decompilation (sc^2dec) method recompiles the LLM's decompilation results to construct pairs for in-context learning, helping the model improve decompilation performance. (2) Fine-grained Alignment Enhancement (FAE), which meticulously aligns assembly code with source code at the statement level by leveraging debugging information, is employed during the fine-tuning phase to achieve further improvements in decompilation. By integrating these two methods, we achieved a Re-Executability performance improvement of approximately 7.35\% on the Decompile-Eval benchmark, establishing a new state-of-the-art performance of 55.03\%.

LLM agents often retry after external validation rejects a candidate, but the interface between validation and the next model call remains underspecified. We introduce VeriHarness, a code-controlled agent loop in which models generate candidates while external validators control acceptance, budgets, and traces. We use it to compare raw diagnostics with feedback that identifies the failure location, observed value, and admissible alternatives. Across 50 paired TextWorld games under a four-call cap, feedback containing all three fields raises terminal success from 14/50 to 36/50 for Qwen2.5-Coder-14B (+44 percentage points) and from 8/50 to 29/50 for Llama-3.1-8B (+42 points). Ablations locate most of the gain in the admissible alternatives: feedback containing only the location and observed value remains near the raw diagnostic baseline. Presenting the complete repair information in prose instead of a keyed JSON record yields nearly the same success, providing no evidence that JSON syntax itself improves repair. The ordering persists across the tested call budgets and one sampled-decoding setting.

Although agent skills equip LLMs with reusable procedural knowledge, manual maintenance suffers from high costs, unscalability, and misalignment. Real-world deployments thus require autonomous, on-demand skill evolution at test time, constrained by limited interaction budgets and a lack of training or validation sets. This setting introduces a severe sparse reward challenge, where outcomes conflate multiple latent failure causes. Under such ambiguity, existing methods that greedily refine a single incumbent skill are particularly vulnerable to an exploitation trap, allowing early misdiagnoses to exhaust limited trials along unproductive trajectories. To address this, we introduce SkillHEX, a closed-loop framework coupling hypothesis-driven self-verification with evidence-guided tree search. SkillHEX translates falsifiable failure hypotheses into executable tests, producing diagnostic evidence as dense reward without additional environment attempts. This evidence guides a search over persistent skill-revision branches, dynamically balancing the exploitation of supported edits with the exploration of plausible alternatives. Evaluated on 87 tasks from SkillsBench, SkillHEX outperforms existing self-evolving methods and achieves an average pass rate of 55.9% and 57.9% using GPT-5.3-Codex and Claude Opus 4.7 under a five-iteration budget, respectively.

A reliable executable environment is the foundation for ensuring that large language models solve software engineering tasks. Due to the complex and tedious construction process, large-scale configuration is relatively inefficient. However, most methods always overlook fine-grained analysis of the actions performed by the agent, making it difficult to handle complex errors and resulting in configuration failures. To address this bottleneck, we propose EvoConfig, an efficient environment configuration framework that optimizes multi-agent collaboration to build correct runtime environments. EvoConfig features an expert diagnosis module for fine-grained post-execution analysis, and a self-evolving mechanism that lets expert agents self-feedback and dynamically adjust error-fixing priorities in real time. Empirically, EvoConfig matches the previous state-of-the-art Repo2Run on Repo2Run's 420 repositories, while delivering clear gains on harder cases: on the more challenging Envbench, EvoConfig achieves a 78.1% success rate, outperforming Repo2Run by 7.1%. Beyond end-to-end success, EvoConfig also demonstrates stronger debugging competence, achieving higher accuracy in error identification and producing more effective repair recommendations than existing methods.

We pursue a vision for self-improving language models in which the model does not merely generate problems or traces to imitate, but constructs the environments that train it. In zero-data reasoning RL, this reframes self-improvement from a data-generation loop into an environment-construction loop, where each artifact is a reusable executable object that samples instances, computes references, and scores responses. Whether this vision sustains improvement hinges on a single property: the environments must exhibit stable solve--verify asymmetry, the model must be able to write an oracle once that it cannot reliably execute in natural language on fresh instances. This asymmetry takes two complementary forms. Some tasks are algorithmically hard to reason through but trivial as code: a dynamic program or graph traversal, compiled once, yields unboundedly many calibrated instances. Others are intrinsically hard to solve but easy to verify, like planted subset-sum or constraint satisfaction. Both create a durable gap between proposing and solving that the policy cannot close by gaming the verifier, and it is this gap that keeps reward informative as the learner improves. We instantiate this view in EvoEnv, a single-policy generator, solver method that synthesizes Python environments from ten seeds and admits them only after staged validation, semantic self-review, solver-relative difficulty calibration, and novelty checks. The strongest evidence comes from the already-strong regime: on Qwen3-4B-Thinking, fixed public-data RLVR and fixed hand-crafted environment RLVR reduce the average, while EvoEnv improves it from 72.4 to 74.8, a relative gain of 3.3%. Stable self-improvement, we suggest, depends not on producing more synthetic data, but on models learning to construct worlds whose difficulty stays structurally beyond their own reach.

Code completion has become an essential tool for daily software development. Existing evaluation benchmarks often employ static methods that do not fully capture the dynamic nature of real-world coding environments and face significant challenges, including limited context length, reliance on superficial evaluation metrics, and potential overfitting to training datasets. In this work, we introduce a novel framework for enhancing code completion in software development through the creation of a repository-level benchmark ExecRepoBench and the instruction corpora Repo-Instruct, aim at improving the functionality of open-source large language models (LLMs) in real-world coding scenarios that involve complex interdependencies across multiple files. ExecRepoBench includes 1.2K samples from active Python repositories. Plus, we present a multi-level grammar-based completion methodology conditioned on the abstract syntax tree to mask code fragments at various logical units (e.g. statements, expressions, and functions). Then, we fine-tune the open-source LLM with 7B parameters on Repo-Instruct to produce a strong code completion baseline model Qwen2.5-Coder-Instruct-C based on the open-source model. Qwen2.5-Coder-Instruct-C is rigorously evaluated against existing benchmarks, including MultiPL-E and ExecRepoBench, which consistently outperforms prior baselines across all programming languages. The deployment of can be used as a high-performance, local service for programming development\url{https://execrepobench.github.io/}.

Large language models (LLMs) are increasingly asked not only to write static interfaces, but to construct executable interactive worlds from natural language. Browser-native 3D, commonly built with Three.js, is a natural next frontier: generated programs must integrate assets, obey spatial and physical constraints, and keep user-facing controls synchronized with hidden runtime state. Existing web-generation benchmarks and evaluators, however, largely observe only pixels or DOM nodes, while the mechanics of a Three.js world unfold inside an opaque <canvas>. We introduce WorldCoder-Bench, a benchmark for autonomous, physically grounded 3D world synthesis. WorldCoder-Bench contains 2,026 expert-curated tasks across Simulation, Rendering, and Application scenarios, with optional .glb assets and hidden behavioral contracts. We further propose StateProbe, an execution-based protocol that probes generated programs in a sandboxed browser and verifies hidden, mutation-hardened contracts over runtime states and transitions. Beyond verification coverage, we report Return on Automation and Time Efficiency Multiplier to measure correctness-adjusted cost and time savings. Across nine frontier models, the best system reaches only 27.8% verification coverage on WorldCoder-Core and 19.9% on WorldCoder-Robust, with failures dominated by state-schema drift and broken interaction chains rather than missing scene elements. Utility metrics further show that cheap or fast models can still provide substantial value on easier domains. WorldCoder-Bench is available at https://anonymous.4open.science/r/WorldCoder-Bench/.

Vision-language process reward models (VL-PRMs) are increasingly used to score intermediate reasoning steps and rerank candidates under test-time scaling. However, they often function as black-box judges: a low step score may reflect a genuine reasoning mistake or simply the verifier's misperception of the image. This entanglement between perception and reasoning leads to systematic false positives (rewarding hallucinated visual premises) and false negatives (penalizing correct grounded statements), undermining both reranking and error localization. We introduce Explicit Visual Premise Verification (EVPV), a lightweight verification interface that conditions step scoring on the reliability of the visual premises a step depends on. The policy is prompted to produce a step-wise visual checklist that makes required visual facts explicit, while a constraint extractor independently derives structured visual constraints from the input image. EVPV matches checklist claims against these constraints to compute a scalar visual reliability signal, and calibrates PRM step rewards via reliability gating: rewards for visually dependent steps are attenuated when reliability is low and preserved when reliability is high. This decouples perceptual uncertainty from logical evaluation without per-step tool calls. Experiments on VisualProcessBench and six multimodal reasoning benchmarks show that EVPV improves step-level verification and consistently boosts Best-of-N reranking accuracy over strong baselines. Furthermore, injecting controlled corruption into the extracted constraints produces monotonic performance degradation, providing causal evidence that the gains arise from constraint fidelity and explicit premise verification rather than incidental prompt effects. Code is available at: https://github.com/Qwen-Applications/EVPV-PRM

Detecting LLM reasoning failures at inference time without ground-truth labels has motivated a wide range of confidence baselines, including self-consistency, semantic entropy, and P(True), built on within-question sampling and self-evaluation. Operad theory, the formalism for systems built by iterated substitution, suggests a complementary diagnostic: a model's direct answer to a compositional query should agree with the answer it produces by composing a stated decomposition of the same query. We instantiate this idea as operadic consistency (OC), a per-question signal. Across twelve instruction-tuned LLMs (4B to 671B parameters, open-weights and closed-source) on four multi-hop QA datasets, OC is strongly correlated with accuracy on every dataset (Pearson r in [0.86, 0.94], all p leq 0.0004), and is the only signal we evaluate with r geq 0.85 uniformly across all four datasets. Chain-of-thought self-consistency (CoT-SC; Wang et al., 2023) matches OC on HotpotQA and DROP (r = 0.93, 0.87) but drops to r approx 0.45 on MuSiQue and StrategyQA. At the per-question level, OC contributes information beyond CoT-SC and semantic entropy on every dataset (cluster-robust p leq 10^{-16} for the OC coefficient), and the conclusion is robust to additionally controlling for constructed decomposition-aware baselines (p leq 10^{-13}). The same signal yields selective-prediction improvements (accuracy at fixed coverage) over a tuned CoT-SC baseline at the equal-cost K = 3 budget (AUARC lifts of +0.086 to +0.096 and AUROC lifts of +0.092 to +0.164; 95% CIs exclude zero on every cell). On five frontier thinking models, where the decomposition is extracted from the model's own chain of thought, the same equal-cost comparison gives positive selective-prediction point-estimate lift on all 16 (dataset, budget, metric) cells tested, with 95% CIs excluding zero on 12 of the 16.

We present ACAR (Adaptive Complexity and Attribution Routing), a measurement framework for studying multi-model orchestration under auditable conditions. ACAR uses self-consistency variance (sigma) computed from N=3 probe samples to route tasks across single-model, two-model, and three-model execution modes. The system is implemented on top of TEAMLLM, a deterministic execution substrate with immutable artifacts and complete decision traces. We evaluate ACAR on 1,510 tasks spanning four benchmarks: MathArena, Reasoning Gym, LiveCodeBench, and SuperGPQA, using Claude Sonnet 4, GPT-4o, and Gemini 2.0 Flash, producing more than 7,550 auditable runs. Results show that sigma-based routing achieves 55.6 percent accuracy, exceeding the two-model baseline of 54.4 percent while avoiding full ensembling on 54.2 percent of tasks. The routing mechanism is model-agnostic and requires no learned components. We also document negative results. First, retrieval augmentation reduced accuracy by 3.4 percentage points, as median retrieval similarity was only 0.167, demonstrating that experience injection without semantic alignment introduces noise rather than grounding. Second, when models agree on incorrect answers (sigma equals zero), no downstream ensemble can recover; this agreement-but-wrong failure mode is intrinsic to self-consistency and bounds achievable accuracy at approximately eight percentage points below full ensembling. Third, attribution estimates based on proxy signals such as response similarity and entropy showed weak correlation with ground-truth leave-one-out values, indicating that practical attribution requires explicit counterfactual computation. This work documents which assumptions fail in practice and provides falsifiable baselines for future research on routing, retrieval, and multi-model attribution.

The automated generation of test code can reduce the time and effort required to build software while increasing its correctness and robustness. In this paper, we present RE-ASSERT, an approach for the automated generation of JUnit test asserts which produces more accurate asserts than previous work with fewer constraints. This is achieved by targeting projects individually, using precise code-to-test traceability for learning and by generating assert statements from the method-under-test directly without the need to write an assert-less test first. We also utilise Reformer, a state-of-the-art deep learning model, along with two models from previous work to evaluate ReAssert and an existing approach, known as ATLAS, using lexical accuracy,uniqueness, and dynamic analysis. Our evaluation of ReAssert shows up to 44% of generated asserts for a single project match exactly with the ground truth, increasing to 51% for generated asserts that compile. We also improve on the ATLAS results through our use of Reformer with 28% of generated asserts matching exactly with the ground truth. Reformer also produces the greatest proportion of unique asserts (71%), giving further evidence that Reformer produces the most useful asserts.

Large language models (LLMs) have demonstrated exceptional capabilities when trained within executable runtime environments, notably excelling at software engineering tasks through verified feedback loops. Yet, scalable and generalizable execution-grounded environments remain scarce, limiting progress in training more capable ML agents. We introduce CTF-Dojo, the first large-scale executable runtime tailored for training LLMs with verifiable feedback, featuring 658 fully functional Capture-The-Flag (CTF)-style challenges containerized in Docker with guaranteed reproducibility. To enable rapid scaling without manual intervention, we develop CTF-Forge, an automated pipeline that transforms publicly available artifacts into ready-to-use execution environments in minutes, eliminating weeks of expert configuration traditionally required. We trained LLM-based agents on just 486 high-quality, execution-verified trajectories from CTF-Dojo, achieving up to 11.6% absolute gains over strong baselines across three competitive benchmarks: InterCode-CTF, NYU CTF Bench, and Cybench. Our best-performing 32B model reaches 31.9% Pass@1, establishing a new open-weight state-of-the-art that rivals frontier models like DeepSeek-V3-0324 and Gemini-2.5-Flash. By framing CTF-style tasks as a benchmark for executable-agent learning, CTF-Dojo demonstrates that execution-grounded training signals are not only effective but pivotal in advancing high-performance ML agents without dependence on costly proprietary systems.

Deployed LLM agents rely on agentic context, the model-external textual control content assembled by an operational harness. In this work, the mutable component of that context is a persistent system-level instruction that is updated from operational experience while the model, tools, and harness remain fixed. Over long evolution horizons, flat-text maintenance makes verification increasingly difficult as accumulated instructions grow and interact. We propose Graph-Regularized Agentic Context Evolution (GRACE), which maintains the persistent instruction component as a typed semantic graph and validates proposed updates within the local typed neighborhoods of modified nodes. Accepted graph updates are reconstructed as incremental edits to the textual instruction checkpoint used at deployment. We evaluate GRACE within a fixed telecom agent harness derived from τ^2-bench under a controlled distribution-shift protocol. Across five independent replications, GRACE improves strict reliability, measured by pass^3, from the Gemini 2.5 Flash zero-shot value of 0.091 to 0.673pm0.136 at the final checkpoint. This exceeds a Gemini 3.1 Pro zero-shot reference of 0.242 on the same held-out set, while the flat-text HCE baseline finishes at 0.191pm0.051. These results identify two requirements for reliable long-horizon context evolution, a structural substrate that makes verification local and a consolidation mechanism that keeps accumulated instruction content usable.

Multimodal automated fact-checking (MAFC) verifies claims by retrieving and reasoning over external evidence. However, most existing static benchmarks risk contamination: they primarily consist of outdated claims verifiable using an LLM's internal knowledge without external evidence. This can inflate performance estimates and fail to reflect true capability on novel claims that require up-to-date information. To address this, emerging dynamic benchmarks collect claims published after LLMs' knowledge cut-off dates, assuming they are uncontaminated. This work revisits this assumption by empirically studying contamination risks in both the state-of-the-art (SOTA) static AVeriTeC benchmark and our newly constructed dynamic ClaimReview2025Q4 benchmark, as well as their impact on MAFC evaluation. Our experiments yield 16 findings, highlighting three key results: (1) Dynamic evaluation reduces but does not eliminate contamination risks, as 17.09\%--29.30\% of post-cut-off claims remain potentially contaminated; (2) Many newly published claims can be verified either directly or by synthesizing multiple pieces of public knowledge available before the cut-off; and (3) Contamination can induce statistically significant inflation in MAFC performance, increasing Macro-F1 by up to 11.34 points and distorting system rankings. In light of these findings, we re-evaluate SOTA LLMs under a strictly contamination-controlled setting. Our study provides practical guidelines for trustworthy MAFC evaluation.

Large language models (LLMs) are trained on extensive datasets that encapsulate substantial world knowledge. However, their outputs often include confidently stated inaccuracies. Earlier works suggest that LLMs encode truthfulness as a distinct linear feature, termed the "truth direction", which can classify truthfulness reliably. We address several open questions about the truth direction: (i) whether LLMs universally exhibit consistent truth directions; (ii) whether sophisticated probing techniques are necessary to identify truth directions; and (iii) how the truth direction generalizes across diverse contexts. Our findings reveal that not all LLMs exhibit consistent truth directions, with stronger representations observed in more capable models, particularly in the context of logical negation. Additionally, we demonstrate that truthfulness probes trained on declarative atomic statements can generalize effectively to logical transformations, question-answering tasks, in-context learning, and external knowledge sources. Finally, we explore the practical application of truthfulness probes in selective question-answering, illustrating their potential to improve user trust in LLM outputs. These results advance our understanding of truth directions and provide new insights into the internal representations of LLM beliefs. Our code is public at https://github.com/colored-dye/truthfulness_probe_generalization

The integration of Large Language Models (LLMs) into Geographic Information Systems (GIS) marks a paradigm shift toward autonomous spatial analysis. However, evaluating these LLM-based agents remains challenging due to the complex, multi-step nature of geospatial workflows. Existing benchmarks primarily rely on static text or code matching, neglecting dynamic runtime feedback and the multimodal nature of spatial outputs. To address this gap, we introduce GeoAgentBench (GABench), a dynamic and interactive evaluation benchmark tailored for tool-augmented GIS agents. GABench provides a realistic execution sandbox integrating 117 atomic GIS tools, encompassing 53 typical spatial analysis tasks across 6 core GIS domains. Recognizing that precise parameter configuration is the primary determinant of execution success in dynamic GIS environments, we designed the Parameter Execution Accuracy (PEA) metric, which utilizes a "Last-Attempt Alignment" strategy to quantify the fidelity of implicit parameter inference. Complementing this, a Vision-Language Model (VLM) based verification is proposed to assess data-spatial accuracy and cartographic style adherence. Furthermore, to address the frequent task failures caused by parameter misalignments and runtime anomalies, we developed a novel agent architecture, Plan-and-React, that mimics expert cognitive workflows by decoupling global orchestration from step-wise reactive execution. Extensive experiments with seven representative LLMs demonstrate that the Plan-and-React paradigm significantly outperforms traditional frameworks, achieving the optimal balance between logical rigor and execution robustness, particularly in multi-step reasoning and error recovery. Our findings highlight current capability boundaries and establish a robust standard for assessing and advancing the next generation of autonomous GeoAI.

Recent benchmarks have probed factual consistency and rhetorical robustness in Large Language Models (LLMs). However, a knowledge gap exists regarding how directional framing of factually true statements influences model agreement, a common scenario for LLM users. AssertBench addresses this by sampling evidence-supported facts from FEVEROUS, a fact verification dataset. For each (evidence-backed) fact, we construct two framing prompts: one where the user claims the statement is factually correct, and another where the user claims it is incorrect. We then record the model's agreement and reasoning. The desired outcome is that the model asserts itself, maintaining consistent truth evaluation across both framings, rather than switching its evaluation to agree with the user. AssertBench isolates framing-induced variability from the model's underlying factual knowledge by stratifying results based on the model's accuracy on the same claims when presented neutrally. In doing so, this benchmark aims to measure an LLM's ability to "stick to its guns" when presented with contradictory user assertions about the same fact. The complete source code is available at https://github.com/achowd32/assert-bench.

Repository-level coding benchmarks such as SWE-bench have driven a rapid surge in the capabilities of coding agents. Yet they usually treat coding tasks as a holistic, binary prediction problem (e.g., resolved or unresolved), neglecting fine-grained agent capabilities such as repository understanding, context retrieval, code localization, and bug diagnosis. In this paper, we introduce SWE-Explore, a benchmark that isolates the evaluation of repository exploration, a critical capability of coding agents. Given a repository and an issue, SWE-Explore asks an explorer to return a ranked list of relevant code regions under a fixed line budget. SWE-Explore covers 848 issues across 10 programming languages and 203 open-source repositories. For each instance, we derive line-level ground truth from independent agent trajectories that successfully solved the same issue, distilling the specific code regions their solution paths actually consulted. We evaluate exploration along coverage, ranking, and context-efficiency dimensions, showing that these metrics strongly track downstream repair behavior. Across a broad set of retrieval methods, general coding agents, and specialized localizers, we find that agentic explorers form a clear tier above classical retrieval. While file-level localization is already strong for modern methods, line-level coverage and efficient ranking remain the key axes differentiating state-of-the-art explorers.

Vision-Language Models (VLMs) often generate plausible but incorrect responses to visual queries. However, reliably quantifying the effect of such hallucinations in free-form responses to open-ended queries is challenging as it requires visually verifying each claim within the response. We propose Programmatic VLM Evaluation (PROVE), a new benchmarking paradigm for evaluating VLM responses to open-ended queries. To construct PROVE, we provide a large language model (LLM) with a high-fidelity scene-graph representation constructed from a hyper-detailed image caption, and prompt it to generate diverse question-answer (QA) pairs, as well as programs that can be executed over the scene graph object to verify each QA pair. We thus construct a benchmark of 10.5k challenging but visually grounded QA pairs. Next, to evaluate free-form model responses to queries in PROVE, we propose a programmatic evaluation strategy that measures both the helpfulness and truthfulness of a response within a unified scene graph-based framework. We benchmark the helpfulness-truthfulness trade-offs of a range of VLMs on PROVE, finding that very few are in-fact able to achieve a good balance between the two. Project page: https://prove-explorer.netlify.app/.

A promising research direction in enabling LLMs to generate consistently correct code involves addressing their inability to properly estimate program execution, particularly for code they generate. In this work, we demonstrate that Code LLMs can be trained to simulate program execution in a step-by-step manner and that this capability can be leveraged to improve competitive programming performance. Our approach combines supervised fine-tuning on natural language execution traces, textual explanations grounded in true execution, with reinforcement learning using verifiable rewards. We introduce two complementary objectives: output prediction given code and inputs, and solving competitive programming tasks with either ground-truth or self-predicted execution feedback. These objectives enable models to perform self-verification over multiple candidate solutions, and iterative self-fixing by simulating test execution. Across multiple competitive programming benchmarks, our method yields consistent improvements over standard reasoning approaches. We further present ablations and analysis to elucidate the role of execution simulation and its limitations.

Tool-calling benchmarks are increasingly used to rank language-model agents, yet their scores are often treated as ground truth without validating the evaluators themselves. We present a systematic validity and reproducibility audit of four major tool-calling benchmark families: BFCL v4, τ2-Bench, LiveMCPBench, and MCP-Atlas. Across 496 expert-reviewed benchmark tasks, we find 92 evaluator-human disagreements, corresponding to an 18.5% misalignment rate. The failures are not isolated annotation mistakes: deterministic benchmarks exhibit brittle state matching, trajectory lock-in, incorrect ground truths, substring-based communication failures, and reward-basis misalignment, while LLM-judge benchmarks exhibit rubric drift, hallucinated completion, answer-only scoring, and substantial run-to-run variance. In LiveMCPBench, 23 repeated evaluations of the same setup produce scores ranging from 57.9% to 76.8%, a spread of 18.9 percentage points, large enough to change leaderboard conclusions. These results show that current tool-calling scores can reflect evaluator artifacts rather than agent capability. We introduce a unified taxonomy of tool-calling evaluation failures, release trace-level audit artifacts and corrected evaluation components, and argue for decomposed metrics that separately measure tool invocation, task completion, and outcome verification. Our findings suggest that progress in tool-using agents requires benchmarks whose evaluators are themselves reproducible, auditable, and aligned with human judgments of task success. We further introduce Tool-Veritas, a configurable benchmark that combines deterministic state verification with optional qualitative judging, and Harness Lab, an open-source system for benchmark execution, trace inspection, repeated-run comparison, and evaluator debugging.

A fundamental skill among human developers is the ability to understand and reason about program execution. As an example, a programmer can mentally simulate code execution in natural language to debug and repair code (aka. rubber duck debugging). However, large language models (LLMs) of code are typically trained on the surface textual form of programs, thus may lack a semantic understanding of how programs execute at run-time. To address this issue, we propose NExT, a method to teach LLMs to inspect the execution traces of programs (variable states of executed lines) and reason about their run-time behavior through chain-of-thought (CoT) rationales. Specifically, NExT uses self-training to bootstrap a synthetic training set of execution-aware rationales that lead to correct task solutions (e.g., fixed programs) without laborious manual annotation. Experiments on program repair tasks based on MBPP and HumanEval demonstrate that NExT improves the fix rate of a PaLM 2 model, by 26.1% and 14.3% absolute, respectively, with significantly improved rationale quality as verified by automated metrics and human raters. Our model can also generalize to scenarios where program traces are absent at test-time.

LLM-based agents have demonstrated strong capabilities in data-intensive analytical tasks, yet their outputs are rarely verifiable: a reliance on linear text trajectories makes their reasoning difficult to audit. In particular, deterministic computations over raw data and semantic deductions over natural-language claims are often entangled in an unstructured stream, leaving numerical conclusions hard to reproduce and qualitative judgments hard to inspect. To address this, we propose VeriGraph, a traceable neuro-symbolic reasoning framework that enables agents to construct an explicit heterogeneous evidence directed acyclic graph (DAG) during execution. VeriGraph introduces three evidence-expansion primitives, namely computational, grounding, and derivational expansion, to connect raw data, interpreter variables, computed results, and natural-language claims in a unified graph. Under this formulation, structural traceability is reduced to graph reachability from raw data sources to terminal claims, while semantic support is measured by claim-level evidence evaluation. To improve graph construction, we further design a graph-based policy optimization strategy with a composite reward that jointly supervises answer correctness, computational integrity, and derivational coherence. Experiments on four benchmarks show that VeriGraph-8B achieves the highest overall score among all baselines. More importantly, VeriGraph produces auditable evidence graphs with substantially stronger claim grounding, achieving a 87.61\% Grounding Rate under our claim-level evidence support evaluation. These results suggest that explicit evidence-graph construction is a promising path toward verifiable data-analytic agents. Our code is available at https://github.com/ignorejjj/VeriGraph.

Context engineering has become central to building reliable AI agents, yet it remains largely unmeasured. Agents do not fail in isolation: their behavior is shaped by the instructions, tools, memory, retrieved knowledge, guardrails, and untrusted inputs accumulated in their context. When this context is weak, agents drift, hallucinate, misuse tools, ignore constraints, become vulnerable to injection, and waste tokens. This paper validates context-engineering quality as an independent leading indicator of agent reliability. We implement the measurement in ProofAgent-Harness, an open-source infrastructure for AI agent evaluation that uses multi-juror, consensus-based scoring. The harness assesses context across seven criteria: role clarity, guardrail coverage, instruction consistency, tool schema quality, grounding sufficiency, injection hardening, and token efficiency. Crucially, the context score is isolated from behavioral metrics and release decisions, enabling a non-circular validation. Through a controlled context-quality study across regulated agent domains, holding frontier LLM agents fixed and varying only their operating context, we show that context-quality criteria consistently predict their corresponding behavioral outcomes. Grounding sufficiency predicts hallucination resistance, guardrail coverage predicts manipulation resistance, instruction consistency predicts instruction following, and tool-schema quality predicts tool use. These findings establish context measurement as a validated preflight signal for agent reliability and position context engineering as an auditable layer of agent evaluation and governance.

Large Language Models show promise for agricultural advisory, yet vanilla models exhibit unsupported recommendations, generic advice lacking specific, actionable detail, and communication styles misaligned with smallholder farmer needs. In high stakes agricultural contexts, where recommendation accuracy has direct consequences for farmer outcomes, these limitations pose challenges for responsible deployment. We present a hybrid LLM architecture that decouples factual retrieval from conversational delivery: supervised fine-tuning with LoRA on expert-curated GOLDEN FACTS (atomic, verified units of agricultural knowledge) optimizes fact recall, while a separate stitching layer transforms retrieved facts into culturally appropriate, safety-aware responses. Our evaluation framework, DG-EVAL, performs atomic fact verification (measuring recall, precision, and contradiction detection) against expert-curated ground truth rather than Wikipedia or retrieved documents. Experiments across multiple model configurations on crops and queries from Bihar, India show that fine-tuning on curated data substantially improves fact recall and F1, while maintaining high relevance. Using a fine-tuned smaller model achieves comparable or better factual quality at a fraction of the cost of frontier models. A stitching layer further improves safety subscores while maintaining high conversational quality. We release the farmerchat-prompts library to enable reproducible development of domain-specific agricultural AI.

Large Language Models (LLMs) pre-trained on code have recently emerged as the dominant approach to program synthesis. However, these models are trained using next-token prediction, which ignores the syntax and semantics of code. We propose RLCF, that further trains a pre-trained LLM via reinforcement learning, using feedback from a grounding function that scores the quality of the code. The grounding function uses (i) compiler-derived feedback on whether the code it generates passes a set of correctness checks; and (ii) feedback from a different LLM that compares the generated code to a reference code. RLCF is model- and language-agnostic. We empirically evaluate it on the MBJP and MathQA tasks for Java. Our experiments show that RLCF raises the odds that an LLM-generated program compiles, is executable, and produces the right output on tests, often allowing LLMs to match the performance of 2x-8x larger LLMs.

We address the problem of runtime trajectory anomaly detection, a critical capability for enabling trustworthy LLM agents. Current safety measures predominantly focus on static input/output filtering. However, we argue that ensuring LLM agents reliability requires auditing the intermediate execution process. In this work, we formulate the task of Trajectory Anomaly Detection. The goal is not merely detection, but precise error localization. This capability is essential for enabling efficient rollback-and-retry. To achieve this, we construct TrajBench, a dataset synthesized via a perturb-and-complete strategy to cover diverse procedural anomalies. Using this benchmark, we investigate the capability of models in process supervision. We observe that general-purpose LLMs, even with zero-shot prompting, struggle to identify and localize these anomalies. This reveals that generalized capabilities do not automatically translate to process reliability. To address this, we propose TrajAD, a specialized verifier trained with fine-grained process supervision. Our approach outperforms baselines, demonstrating that specialized supervision is essential for building trustworthy agents.

Background manipulation is a practical but under-specified image-forensics setting: the manipulated evidence can sit outside the salient foreground object, while many evaluations emphasize object-centric copy-move, splicing, or generic synthetic edits. We introduce BG-REAL, a public real-data anchored benchmark package for background manipulation detection and localization. The current release is built from Open Images V7 instance-segmentation sources and contains 7,000 processed samples over 1,200 source groups, including 6,000 public-data anchored samples and 1,000 synthetic control samples. BG-REAL covers six edit families, matched authentic controls, source-group splits, mask and leakage QA, 599 human-assisted quality-control rows, three completed external baselines (TruFor, MVSS-Net, and HiFi-Net), and five-seed model evaluation. Beyond aggregate accuracy, we use matched-authentic-control diagnostics to measure how often baselines misclassify re-encoded authentic images as manipulated at a threshold fixed on held-out validation data; false-positive rates range from 0.57 (TruFor, the lowest) to 1.00 (several weak or mask-informed baselines), indicating that re-encoding artifacts are a shared shortcut risk across baselines rather than a problem specific to any one model. The release provides the construction pipeline, evaluation protocol, paper-ready figures, and reproduction documentation. We frame BG-REAL as a background-manipulation-focused complement to general image-manipulation-localization benchmarks, not as a fully real-only or general-purpose benchmark.

Adding procedural skills to an LLM agent is typically evaluated by average improvement in task success. However, this metric hides an important cost: skills can also make agents worse. We measure both sides by comparing agents with and without skills across nearly 6,000 runs spanning two office automation benchmarks and three model harness stacks. This allows us to distinguish two outcomes. A regression is a task solved without skills but failed after skills are added. A residual failure is a task that fails both with and without skills. We find that regressions are substantial enough that the best performing skills outperform others primarily by regressing less, not by gaining more. We identify three causes of regression: (i) skill description osmosis, a skill changes an agent's behavior simply by being present in context, even when it is never invoked; (ii) grounding displacement, a skill's prescribed procedure overrides how the agent interprets its inputs; and (iii) verification displacement, where the procedure suppresses checks the agent would otherwise perform on its outputs. Analysing persistent failures reveals the same underlying pattern. Existing skills overemphasize procedural guidance the stage least often responsible for failure while under supporting grounding and verification, the dominant sources of remaining errors. After correcting evaluation artifacts and studying traces, we find many regressions and persistent failures recoverable through better grounding and verification. Procedural skills should be evaluated by decomposing their net effect into gains and regressions, not by aggregate improvement alone. We identify three regression modes skills should avoid, and find that reliability depends more on grounding and verification than on procedural skill choice.