Automated Evaluation

Artifact Evaluation (AE) is essential for ensuring the transparency and reliability of research, closing the gap between exploratory work and real-world deployment is particularly important in cybersecurity, particularly in IoT and CPSs, where large-scale, heterogeneous, and privacy-sensitive data meet safety-critical actuation. Yet, manual reproducibility checks are time-consuming and do not scale with growing submission volumes. In this work, we demonstrate that Large Language Models (LLMs) can provide powerful support for AE tasks: (i) text-based reproducibility rating, (ii) autonomous sandboxed execution environment preparation, and (iii) assessment of methodological pitfalls. Our reproducibility-assessment toolkit yields an accuracy of over 72% and autonomously sets up execution environments for 28% of runnable cybersecurity artifacts. Our automated pitfall assessment detects seven prevalent pitfalls with high accuracy ($F_1$ > 92%). Hence, the toolkit significantly reduces reviewer effort and, when integrated into established AE processes, could incentivize authors to submit higher-quality and more reproducible artifacts. IoT, CPS, and cybersecurity conferences and workshops may integrate the toolkit into their peer-review processes to support reviewers' decisions on awarding artifact badges, improving the overall sustainability of the process.

Product concept evaluation is a critical stage that determines strategic resource allocation and project success in enterprises. However, traditional expert-led approaches face limitations such as subjective bias and high time and cost requirements. To support this process, this study proposes an automated approach utilizing a large language model (LLM)-based multi-agent system (MAS). Through a systematic analysis of previous research on product development and team collaboration, this study established two primary evaluation dimensions, namely technical feasibility and market feasibility. The proposed system consists of a team of eight virtual agents representing specialized domains such as R&D and marketing. These agents use retrieval-augmented generation (RAG) and real-time search tools to gather objective evidence and validate concepts through structured deliberations based on the established criteria. The agents were further fine-tuned using professional product review data to enhance their judgment accuracy. A case study involving professional display monitor concepts demonstrated that the system's evaluation rankings were consistent with those of senior industry experts. These results confirm the usability of the proposed multi-agent-based evaluation approach for supporting product development decisions.

Code generation has emerged as one of AI's highest-impact use cases, yet existing benchmarks measure isolated tasks rather than the complete "zero-to-one" process of building a working application from scratch. We introduce Vibe Code Bench, a benchmark of 100 web application specifications (50 private validation, 50 held-out test) with 964 browser-based workflows comprising 10,131 substeps, evaluated against deployed applications by an autonomous browser agent. Across 16 frontier models, the best achieves 61.8% accuracy on the test split, revealing that reliable end-to-end application development remains a frontier challenge. We identify self-testing during generation as a strong performance predictor (Pearson r=0.72), and show through a completed human alignment study that evaluator selection materially affects outcomes (31.8-93.6% pairwise step-level agreement). Our contributions include (1) a novel benchmark dataset and browser-based evaluation pipeline for end-to-end web application development, (2) a comprehensive evaluation of 16 frontier models with cost, latency, and error analysis, and (3) an evaluator alignment protocol with both cross-model and human annotation results.

We present SpotIt+, an open-source tool for evaluating Text-to-SQL systems via bounded equivalence verification. Given a generated SQL query and the ground truth, SpotIt+ actively searches for database instances that differentiate the two queries. To ensure that the generated counterexamples reflect practically relevant discrepancies, we introduce a best-effort constraint-mining pipeline that combines rule-based specification mining with LLM-based validation over example databases. Experimental results on the BIRD dataset show that the mined constraints enable SpotIt+ to generate more realistic differentiating databases, while preserving its ability to efficiently uncover numerous discrepancies between generated and gold SQL queries that are missed by standard test-based evaluation.

Moving beyond evaluations that collapse performance across heterogeneous prompts toward fine-grained evaluation at the prompt level, or within relatively homogeneous subsets, is necessary to diagnose generative models' strengths and weaknesses. Such fine-grained evaluations, however, suffer from a data bottleneck: human gold-standard labels are too costly at this scale, while automated ratings are often misaligned with human judgment. To resolve this challenge, we propose a novel statistical model based on tensor factorization that merges cheap autorater data with a limited set of human gold-standard labels. Specifically, our approach uses autorater scores to pretrain latent representations of prompts and generative models, and then aligns those pretrained representations to human preferences using a small calibration set. This sample-efficient methodology is robust to autorater quality, more accurately predicts human preferences on a per-prompt basis than standard baselines, and provides tight confidence intervals for key statistical parameters of interest. We also showcase the practical utility of our method by constructing granular leaderboards based on prompt qualities and by estimating model performance solely from autorater scores, eliminating the need for additional human annotations.

Topic models uncover latent thematic structures in text corpora, yet evaluating their quality remains challenging, particularly in specialized domains. Existing methods often rely on automated metrics like topic coherence and diversity, which may not fully align with human judgment. Human evaluation tasks, such as word intrusion, provide valuable insights but are costly and primarily validated on general-domain corpora. This paper introduces Topic Word Mixing (TWM), a novel human evaluation task assessing inter-topic distinctness by testing whether annotators can distinguish between word sets from single or mixed topics. TWM complements word intrusion's focus on intra-topic coherence and provides a human-grounded counterpart to diversity metrics. We evaluate six topic models - both statistical and embedding-based (LDA, NMF, Top2Vec, BERTopic, CFMF, CFMF-emb) - comparing automated metrics with human evaluation methods based on nearly 4,000 annotations from a domain-specific corpus of philosophy of science publications. Our findings reveal that word intrusion and coherence metrics do not always align, particularly in specialized domains, and that TWM captures human-perceived distinctness while appearing to align with diversity metrics. We release the annotated dataset and task generation code. This work highlights the need for evaluation frameworks bridging automated and human assessments, particularly for domain-specific corpora.

Many readers today struggle to assess the trustworthiness of online news because reliable reporting coexists with misinformation. The TREC 2025 DRAGUN (Detection, Retrieval, and Augmented Generation for Understanding News) Track provided a venue for researchers to develop and evaluate assistive RAG systems that support readers' news trustworthiness assessment by producing reader-oriented, well-attributed reports. As the organizers of the DRAGUN track, we describe the resources that we have newly developed to allow for the reuse of the track's tasks. The track had two tasks: (Task 1) Question Generation, producing 10 ranked investigative questions; and (Task 2, the main task) Report Generation, producing a 250-word report grounded in the MS MARCO V2.1 Segmented Corpus. As part of the track's evaluation, we had TREC assessors create importance-weighted rubrics of questions with expected short answers for 30 different news articles. These rubrics represent the information that assessors believe is important for readers to assess an article's trustworthiness. The assessors then used their rubrics to manually judge the participating teams' submitted runs. To make these tasks and their rubrics reusable, we have created an automated process to judge runs not part of the original assessing. We show that our AutoJudge ranks existing runs well compared to the TREC human-assessed evaluation (Kendall's $τ= 0.678$ for Task 1 and $τ= 0.872$ for Task 2). These resources enable both the evaluation of RAG systems for assistive news trustworthiness assessment and, with the human evaluation as a benchmark, research on improving automated RAG evaluation.

The pursuit of human-like conversational agents has long been guided by the Turing test. For modern speech-to-speech (S2S) systems, a critical yet unanswered question is whether they can converse like humans. To tackle this, we conduct the first Turing test for S2S systems, collecting 2,968 human judgments on dialogues between 9 state-of-the-art S2S systems and 28 human participants. Our results deliver a clear finding: no existing evaluated S2S system passes the test, revealing a significant gap in human-likeness. To diagnose this failure, we develop a fine-grained taxonomy of 18 human-likeness dimensions and crowd-annotate our collected dialogues accordingly. Our analysis shows that the bottleneck is not semantic understanding but stems from paralinguistic features, emotional expressivity, and conversational persona. Furthermore, we find that off-the-shelf AI models perform unreliably as Turing test judges. In response, we propose an interpretable model that leverages the fine-grained human-likeness ratings and delivers accurate and transparent human-vs-machine discrimination, offering a powerful tool for automatic human-likeness evaluation. Our work establishes the first human-likeness evaluation for S2S systems and moves beyond binary outcomes to enable detailed diagnostic insights, paving the way for human-like improvements in conversational AI systems.

This paper presents a comprehensive evaluation framework for assessing the cultural competence of large language models (LLMs) in Persian. Existing Persian cultural benchmarks rely predominantly on multiple-choice formats and English-centric metrics that fail to capture Persian's morphological complexity and semantic nuance. Our framework introduces a Persian-specific short-answer evaluation that combines rule-based morphological normalization with a hybrid syntactic and semantic similarity module, enabling robust soft-match scoring beyond exact string overlap. Through systematic evaluation of 15 state-of-the-art open- and closed-source models across three culturally grounded Persian datasets, we demonstrate that our hybrid evaluation improves scoring consistency by +10 compared to exact-match baselines by capturing meaning that surface-level methods cannot detect. Our human evaluation further confirms that the proposed semantic similarity metric achieves higher agreement with human judgments than LLM-based judges. We publicly release our evaluation framework, providing the first standardized benchmark for measuring cultural understanding in Persian and establishing a reproducible foundation for cross-cultural LLM evaluation research.

Generative AI is widely used to create commercial posters. However, rapid advances in generation have outpaced automated quality assessment. Existing models emphasize generic esthetics or low level distortions and lack the functional criteria required for e-commerce design. It is especially challenging for Chinese content, where complex characters often produce subtle but critical textual artifacts that are overlooked by existing methods. To address this, we introduce E-comIQ-ZH, a framework for evaluating Chinese e-commerce posters. We build the first dataset E-comIQ-18k to feature multi dimensional scores and expert calibrated Chain of Thought (CoT) rationales. Using this dataset, we train E-comIQ-M, a specialized evaluation model that aligns with human expert judgment. Our framework enables E-comIQ-Bench, the first automated and scalable benchmark for the generation of Chinese e-commerce posters. Extensive experiments show our E-comIQ-M aligns more closely with expert standards and enables scalable automated assessment of e-commerce posters. All datasets, models, and evaluation tools will be released to support future research in this area.Code will be available at https://github.com/4mm7/E-comIQ-ZH.

Maqam, a singing type, is a significant component of Kurdish music. A maqam singer receives training in a traditional face-to-face or through self-training. Automatic Singing Assessment (ASA) uses machine learning (ML) to provide the accuracy of singing styles and can help learners to improve their performance through error detection. Currently, the available ASA tools follow Western music rules. The musical composition requires all notes to stay within their expected pitch range from start to finish. The system fails to detect micro-intervals and pitch bends, so it identifies Kurdish maqam singing as incorrect even though the singer performs according to traditional rules. Kurdish maqam requires recognizing performance errors within microtonal spaces, which is beyond Western equal temperament. This research is the first attempt to address the mentioned gap. While many error types happen during singing, our focus is on pitch, rhythm, and modal stability errors in the context of Bayati-Kurd. We collected 50 songs from 13 vocalists ( 2-3 hours) and annotated 221 error spans (150 fine pitch, 46 rhythm, 25 modal drift). The data was segmented into 15,199 overlapping windows and converted to log-mel spectrograms. We developed a two-headed CNN-BiLSTM with attention mode to decide whether a window contains an error and to classify it based on the chosen errors. Trained for 20 epochs with early stopping at epoch 10, the model reached a validation macro-F1 of 0.468. On the full 50-song evaluation at a 0.750 threshold, recall was 39.4% and precision 25.8% . Within detected windows, type macro-F1 was 0.387, with F1 of 0.492 (fine pitch), 0.536 (rhythm), and 0.133 (modal drift); modal drift recall was 8.0%. The better performance on common error types shows that the method works, while the poor modal-drift recall shows that more data and balancing are needed.

Comparative evaluation of several systems is a recurrent task in researching. It is a key step before deciding which system to use for our work, or, once our research has been conducted, to demonstrate the potential of the resulting model. Furthermore, it is the main task of competitive, public challenges evaluation. Our proposed software (DEEP) automates both the execution and scoring of machine translation and optical character recognition models. Furthermore, it is easily extensible to other tasks. DEEP is prepared to receive dockerized systems, run them (extracting information at that same time), and assess hypothesis against some references. With this approach, evaluators can achieve a better understanding of the performance of each model. Moreover, the software uses a clustering algorithm based on a statistical analysis of the significance of the results yielded by each model, according to the evaluation metrics. As a result, evaluators are able to identify clusters of performance among the swarm of proposals and have a better understanding of the significance of their differences. Additionally, we offer a visualization web-app to ensure that the results can be adequately understood and interpreted. Finally, we present an exemplary case of use of DEEP.

Advances in large language models (LLMs) are rapidly transforming scientific work, yet empirical evidence on how these systems reshape research activities remains limited. We report a mixed-methods pilot evaluation of an AI-orchestrated research workflow in which a human researcher coordinated multiple LLM-based agents to perform data extraction, corpus construction, artifact generation, and artifact evaluation. Using the generation and assessment of multiple-choice questions (MCQs) as a testbed, we collected 1,071 SAT Math MCQs and employed LLM agents to extract questions from PDFs, retrieve and convert open textbooks into structured representations, align each MCQ with relevant textbook content, generate new MCQs under specified difficulty and cognitive levels, and evaluate both original and generated MCQs using a 24-criterion quality framework. Across all evaluations, average MCQ quality was high. However, criterion-level analysis and equivalence testing show that generated MCQs are not fully comparable to expert-vetted baseline questions. Strict similarity (24/24 criteria equivalent) was never achieved. Persistent gaps concentrated in skill\ depth, cognitive engagement, difficulty calibration, and metadata alignment, while surface-level qualities, such as {grammar fluency}, {clarity options}, {no duplicates}, were consistently strong. Beyond MCQ outcomes, the study documents a labor shift. The researcher's work moved from authoring items'' toward {specification, orchestration, verification}, and {governance}. Formalizing constraints, designing rubrics, building validation loops, recovering from tool failures, and auditing provenance constituted the primary activities. We discuss implications for the future of scientific work, including emerging AI research operations'' skills required for AI-empowered research pipelines.

Robust and comprehensive evaluation of large language models (LLMs) is essential for identifying effective LLM system configurations and mitigating risks associated with deploying LLMs in sensitive domains. However, traditional statistical metrics are poorly suited to open-ended generation tasks, leading to growing reliance on LLM-based evaluation methods. These methods, while often more flexible, introduce additional complexity: they depend on carefully chosen models, prompts, parameters, and evaluation strategies, making the evaluation process prone to misconfiguration and bias. In this work, we present EvalSense, a flexible, extensible framework for constructing domain-specific evaluation suites for LLMs. EvalSense provides out-of-the-box support for a broad range of model providers and evaluation strategies, and assists users in selecting and deploying suitable evaluation methods for their specific use-cases. This is achieved through two unique components: (1) an interactive guide aiding users in evaluation method selection and (2) automated meta-evaluation tools that assess the reliability of different evaluation approaches using perturbed data. We demonstrate the effectiveness of EvalSense in a case study involving the generation of clinical notes from unstructured doctor-patient dialogues, using a popular open dataset. All code, documentation, and assets associated with EvalSense are open-source and publicly available at https://github.com/nhsengland/evalsense.

While Multimodal Large Language Models (MLLMs) perform strongly on single-turn chart generation, their ability to support real-world exploratory data analysis remains underexplored. In practice, users iteratively refine visualizations through multi-turn interactions that require maintaining common ground, tracking prior edits, and adapting to evolving preferences. We introduce ChartEditBench, a benchmark for incremental, visually grounded chart editing via code, comprising 5,000 difficulty-controlled modification chains and a rigorously human-verified subset. Unlike prior one-shot benchmarks, ChartEditBench evaluates sustained, context-aware editing. We further propose a robust evaluation framework that mitigates limitations of LLM-as-a-Judge metrics by integrating execution-based fidelity checks, pixel-level visual similarity, and logical code verification. Experiments with state-of-the-art MLLMs reveal substantial degradation in multi-turn settings due to error accumulation and breakdowns in shared context, with strong performance on stylistic edits but frequent execution failures on data-centric transformations. ChartEditBench, establishes a challenging testbed for grounded, intent-aware multimodal programming.

The rapid evolution of Large Language Models has catalyzed a surge in scientific idea production, yet this leap has not been accompanied by a matching advance in idea evaluation. The fundamental nature of scientific evaluation needs knowledgeable grounding, collective deliberation, and multi-criteria decision-making. However, existing idea evaluation methods often suffer from narrow knowledge horizons, flattened evaluation dimensions, and the inherent bias in LLM-as-a-Judge. To address these, we regard idea evaluation as a knowledge-grounded, multi-perspective reasoning problem and introduce InnoEval, a deep innovation evaluation framework designed to emulate human-level idea assessment. We apply a heterogeneous deep knowledge search engine that retrieves and grounds dynamic evidence from diverse online sources. We further achieve review consensus with an innovation review board containing reviewers with distinct academic backgrounds, enabling a multi-dimensional decoupled evaluation across multiple metrics. We construct comprehensive datasets derived from authoritative peer-reviewed submissions to benchmark InnoEval. Experiments demonstrate that InnoEval can consistently outperform baselines in point-wise, pair-wise, and group-wise evaluation tasks, exhibiting judgment patterns and consensus highly aligned with human experts.

Vision-Language Models (VLMs) are increasingly used in document processing pipelines to convert flowchart images into structured code (e.g., Mermaid). In production, these systems process arbitrary inputs for which no ground-truth code exists, making output quality difficult to assess. We propose a reference-free evaluation framework that monitors flowchart image-to-code generation quality at inference time, using only the input image and the generated output. The framework introduces two automated metrics: $\text{Recall}{\text{OCR}}$, which estimates content coverage by extracting text from the input image via OCR as a proxy reference, and $\text{Precision}{\text{VE}}$, which detects hallucinated elements through Visual Entailment against the original image. Their harmonic mean, $\text{F1}{\text{OCR-VE}}$, provides a unified quality score. Validation on the FlowVQA dataset shows strong agreement with ground-truth metrics (average Pearson's $r = 0.97$, $0.91$, and $0.94$ for Recall, Precision, and F1, respectively), confirming the framework's reliability as a practical, reference-free alternative for continuous quality monitoring in production settings.

Legal text generated by large language models (LLMs) can usually achieve reasonable factual accuracy, but it frequently fails to adhere to the specialised stylistic norms and linguistic conventions of legal writing. In order to improve stylistic quality, a crucial first step is to establish a reliable evaluation method. However, having legal experts manually develop such a metric is impractical, as the implicit stylistic requirements in legal writing practice are difficult to formalise into explicit rubrics. Meanwhile, existing automatic evaluation methods also fall short: reference-based metrics conflate semantic accuracy with stylistic fidelity, and LLM-as-a-judge evaluations suffer from opacity and inconsistency. To address these challenges, we introduce CLASE (Chinese LegAlese Stylistic Evaluation), a hybrid evaluation method that focuses on the stylistic performance of legal text. The method incorporates a hybrid scoring mechanism that combines 1) linguistic feature-based scores and 2) experience-guided LLM-as-a-judge scores. Both the feature coefficients and the LLM scoring experiences are learned from contrastive pairs of authentic legal documents and their LLM-restored counterparts. This hybrid design captures both surface-level features and implicit stylistic norms in a transparent, reference-free manner. Experiments on 200 Chinese legal documents show that CLASE achieves substantially higher alignment with human judgments than traditional metrics and pure LLM-as-a-judge methods. Beyond improved alignment, CLASE provides interpretable score breakdowns and suggestions for improvements, offering a scalable and practical solution for professional stylistic evaluation in legal text generation (Code and data for CLASE is available at: https://github.com/rexera/CLASE).

Millions of people now use generative AI chatbots for psychological support. Despite their promise, the most pressing question in AI for mental health is whether these tools are safe. The field currently lacks a validated, automated benchmark for evaluating AI chatbot safety, particularly for users at risk of suicide. The Validation of Ethical and Responsible AI in Mental Health (VERA-MH) evaluation was recently proposed to address this need. This human validation study examined the alignment of VERA-MH safety ratings with expert clinician judgments. We simulated conversations between large language model (LLM)-based users spanning a range of suicide risk levels and disclosure styles and general-purpose AI chatbots. Licensed mental health clinicians from Spring Health independently rated chatbot safety using the VERA-MH scoring rubric. An LLM-based evaluator ("judge") applied the same rubric to the same conversations. We examined agreement among clinicians, between clinician consensus and the LLM judge, and across different judge LLMs. Clinicians also rated user-agent realism, suicide risk, and disclosure. Clinicians showed strong agreement in safety ratings (chance-corrected inter-rater reliability [IRR] = 0.77), establishing a reliable clinical consensus reference. The LLM judge was strongly aligned with this consensus (IRR = 0.81), and ratings were stable across judge models and repeated evaluations. Ratings of user-agent realism and fidelity to intended suicide risk and disclosure styles were mixed. These findings support the reliability of VERA-MH as an open-source, fully automated benchmark for evaluating AI chatbot suicide risk detection and response. Because these results reflect an earlier version of the benchmark, future work should validate updated versions, assess generalizability and robustness, and expand VERA-MH to additional domains of AI safety in mental health.

Users increasingly rely on Large Language Models (LLMs) for Deep Research, using them to synthesize diverse sources into structured reports that support understanding and action. In this context, the practical reliability of such reports hinges on logical quality: whether the report's claims and arguments are explicitly supported and can be trusted as a basis for downstream use, rather than merely appearing fluent or informative. However, current evaluation frameworks largely overlook this requirement. To bridge this gap, we introduce ReportLogic, a benchmark that quantifies report-level logical quality through a reader-centric lens of auditability. Specifically, ReportLogic adopts a hierarchical taxonomy that evaluates whether readers can (1) trace an on-topic report structure with a unified analytical arc (Macro-Logic), (2) understand the progression with necessary context (Expositional-Logic), and (3) verify conclusions via explicit claim--support (Structural-Logic). Based on this taxonomy, we construct a human-annotated rubric-guided dataset and train an open-source LogicJudge for scalable evaluation. We further evaluate judge robustness via adversarial attacks, showing that off-the-shelf LLM judges are frequently influenced by superficial cues (e.g., verbosity), and reasoning modes can mask broken support relations. Overall, our results provide actionable guidance for building more robust logic evaluators and improving the logical reliability of LLM-generated reports.

Accurate evaluation of user satisfaction is critical for iterative development of conversational AI. However, for open-ended assistants, traditional A/B testing lacks reliable metrics: explicit feedback is sparse, while implicit metrics are ambiguous. To bridge this gap, we introduce BoRP (Bootstrapped Regression Probing), a scalable framework for high-fidelity satisfaction evaluation. Unlike generative approaches, BoRP leverages the geometric properties of LLM latent space. It employs a polarization-index-based bootstrapping mechanism to automate rubric generation and utilizes Partial Least Squares (PLS) to map hidden states to continuous scores. Experiments on industrial datasets show that BoRP (Qwen3-8B/14B) significantly outperforms generative baselines (even Qwen3-Max) in alignment with human judgments. Furthermore, BoRP reduces inference costs by orders of magnitude, enabling full-scale monitoring and highly sensitive A/B testing via CUPED.

Evaluating conversational systems in multi-turn settings remains a fundamental challenge. Conventional pipelines typically rely on manually defined rubrics and fixed conversational context$-$a static approach that limits coverage and fails to capture the diverse, emergent behaviors of dialogue models. To address this, we introduce CoReflect (A Reflective Co-Evolution Framework for Improving Conversational Evaluation), which unifies dialogue simulation and evaluation into an adaptive, iterative process. CoReflect employs a conversation planner that generates structured templates to guide a user simulator through diverse, goal-directed dialogues. Subsequently, a reflective analyzer processes these dialogues to identify systematic behavioral patterns and automatically refine the evaluation rubrics. Crucially, the insights from the conversation analysis are fed back into the planner to update conversation templates for subsequent iterations. This co-evolution loop ensures that the complexity of test cases and the diagnostic precision of rubrics improve in tandem. By minimizing human intervention, CoReflect provides a scalable and self-refining methodology that allows evaluation protocols to adapt alongside the rapidly advancing capabilities of dialogue models.

As Large Language Models (LLMs) increasingly operate as Deep Research (DR) Agents capable of autonomous investigation and information synthesis, reliable evaluation of their task performance has become a critical bottleneck. Current benchmarks predominantly rely on static datasets, which suffer from several limitations: limited task generality, temporal misalignment, and data contamination. To address these, we introduce DR-Arena, a fully automated evaluation framework that pushes DR agents to their capability limits through dynamic investigation. DR-Arena constructs real-time Information Trees from fresh web trends to ensure the evaluation rubric is synchronized with the live world state, and employs an automated Examiner to generate structured tasks testing two orthogonal capabilities: Deep reasoning and Wide coverage. DR-Arena further adopts Adaptive Evolvement Loop, a state-machine controller that dynamically escalates task complexity based on real-time performance, demanding deeper deduction or wider aggregation until a decisive capability boundary emerges. Experiments with six advanced DR agents demonstrate that DR-Arena achieves a Spearman correlation of 0.94 with the LMSYS Search Arena leaderboard. This represents the state-of-the-art alignment with human preferences without any manual efforts, validating DR-Arena as a reliable alternative for costly human adjudication.

Deep Research Agents (DRA) aim to help users search the web, synthesize information, and deliver comprehensive investigative reports. Prior benchmarks often either under-evaluate a system's ability to produce meaningful insights and high-quality writing, or adopt coarse or LLM-defined criteria that are hard to verify and can diverge from human expert judgment. To address these issues, we introduce Deep Research Bench II, a new benchmark for evaluating DRAs. It contains 132 grounded research tasks across 22 domains; for each task, an agent must produce a research report that is evaluated by a set of 9,430 fine-grained binary rubrics in total, covering three dimensions: information recall, analysis, and presentation. All rubrics are derived from carefully selected expert-written investigative articles and are constructed through a four-stage LLM+human pipeline that combines automatic extraction with over 400 human-hours of expert review, ensuring that the criteria are verifiable and aligned with human expert judgment. We evaluate several state-of-the-art deep-research agents on Deep Research Bench II and find that even the strongest models satisfy fewer than 50% of the rubrics, revealing a substantial gap between current DRAs and human experts. We release the benchmark, evaluation scripts, and all rubrics at https://github.com/imlrz/DeepResearch-Bench-II to facilitate future research on deep-rearch agents.

As Natural Language Generation (NLG) dominates modern NLP, scalable evaluation remains a critical bottleneck. Consequently, LLM-as-a-judge (LaaJ) adoption has accelerated rapidly, appearing in more papers than human evaluation in 2025. This pivotal shift motivates a critical analysis of current evaluation practices. Overcoming the limits of rigid keyword filtering and manual review, we employ a multi-LLM information extraction pipeline to gather structured metadata from 14,171 papers across four major NLP conferences (2020-2025). Analyzing 3,334 filtered NLG papers, we identify three systemic challenges. (1) Metric inertia: despite the shift toward open-ended generation, legacy lexical metrics (BLEU, ROUGE) persist as primary indicators, typically used alongside rather than replaced by semantic alternatives. (2) Metric-criteria mapping problem: our paper-level co-occurrence data reveals that general-purpose automatic metrics are applied as broad proxies for quality, without specifying which dimension of text generation they are intended to evaluate. (3) Validation gap: LaaJ has grown rapidly without commensurate human validation (fewer than 8% of papers). Crucially, while LaaJ correlates with aggregate quality, alignment collapses on fine-grained criteria like fluency. To address these gaps, we distill our findings into a minimal Evaluation Checklist to guide metric selection, construct validity, and LaaJ deployment.

Applying well-established statistical methods effectively to LLM evals requires consideration of model correlations and prediction noises. Leading evals often ignore data noise, have low statistical power, or make unnecessary trade-offs. We define and measure three types of noise: 1) prediction noise from generating different answers on a given question, 2) data noise from sampling questions, and 3) their combined total noise following the law of total variance. To emphasize relative comparisons and gain statistical power, we propose all-pairs comparison, which applies paired comparison to all pairs of LLMs and measures all the noise components based on millions of question-level predictions across many evals and settings, revealing clear trends. First, most evals exhibit a characteristic and highly predictable total noise level across all model pairs. Second, paired prediction noise typically exceeds paired data noise, which means reducing prediction noise by averaging can significantly increase statistical power. By measuring all the noises together, we can assess eval results in context and make better empirical decisions.

Reliable assessment of the abilities of large audio language models (LALMs) is essential to advancing the state of the art. As benchmarks rapidly evolve to incorporate complex reasoning and subjective tasks, they increasingly necessitate open-ended responses from LALMs. We present Open-ended Response Correctness Assessment (ORCA) -- a reliable and lightweight model-based approach for answer correctness and disagreement modeling. We employ a three-stage annotation pipeline combining human judgment, structured feedback, and human-AI correction, yielding 9,663 annotations across 3,699 question-answer pairs from 15 LALMs on three audio understanding and reasoning benchmarks (achieving a Krippendorff's alpha of 0.82). Our experiments employing curriculum learning show that ORCA models achieve a Spearman correlation of 0.91 with average human correctness ratings on seen benchmarks and generalize to unseen benchmarks with a score of 0.85, outperforming several LLM judge baselines including Gemini 2.5 Flash. Furthermore, we demonstrate that ORCA's predicted variance correlates strongly with human disagreement, allowing it to effectively identify problematic benchmark items.

As LLMs rapidly saturate existing benchmarks, automated benchmark creation using LLMs (LLM as a benchmark) where a model generates test inputs (LLM as a testset) and evaluates outputs (LLM as an evaluator) has gained traction as a cheap alternative to human curation. We show that this paradigm has a fundamental problem: LLM-generated benchmarks systematically favor the model that created them. Using machine translation as our primary testbed, we find that self bias arises from two compounding sources, LLM as a testset and LLM as an evaluator, and their combination amplifies the effect. Crucially, even when test data is generated with explicit diversity controls, each modelś implicit stylistic tendencies produce homogeneous, model-specific outputs that inflate its own scores. Increasing source text diversity, using our proposed diversity metric, partially mitigates this bias. Self bias is strong enough to cause each model to rank itself first, overriding the peer consensus ordering. We confirm that the phenomenon extends to open-ended generation on the Chatbot Arena task.

With the surge of large language models (LLMs) and their ability to produce customized output, style-personalized text generation--"write like me"--has become a rapidly growing area of interest. However, style personalization is highly specific, relative to every user, and depends strongly on the pragmatic context, which makes it uniquely challenging. Although prior research has introduced benchmarks and metrics for this area, they tend to be non-standardized and have known limitations (e.g., poor correlation with human subjects). LLMs have been found to not capture author-specific style well, it follows that the metrics themselves must be scrutinized carefully. In this work we critically examine the effectiveness of the most common metrics used in the field, such as BLEU, embeddings, and LLMs-as-judges. We evaluate these metrics using our proposed style discrimination benchmark, which spans eight diverse writing tasks across three evaluation settings: domain discrimination, authorship attribution, and LLM-generated personalized vs non-personalized discrimination. We find strong evidence that employing ensembles of diverse evaluation metrics consistently outperforms single-evaluator methods, and conclude by providing guidance on how to reliably assess style-personalized text generation.

We evaluate artificial intelligence (AI) systems without ground truth by exploiting a link between strategic gaming and information loss. Building on established information theory, we analyze which mechanisms resist adversarial manipulation. This motivates mutual evaluation, where the overseer is treated as a strategic player estimating mutual information by prompting, making truthful agent reporting an optimal strategy. We show that certain f-divergences, such as total variation distance (TVD), maintain polynomial guarantees under attack, building on an established exponential barrier for estimating mutual information (MI) in worst-case certification settings. Under adversarial attacks, TVD-MI maintains effectiveness (area under the curve 0.70--0.77) while other approaches can decay toward chance, demonstrating that prompting the same system for information relationships rather than quality judgments can improve robustness. The mechanisms decompose pairwise evaluations into reliable item-level detection scores without ground truth, addressing a key limitation of standard peer prediction. Pre-registration: https://osf.io/c7pum .