Responses

Large language models are increasingly used in healthcare communication, yet most evaluations emphasize response quality while assuming that the user's concern has been interpreted correctly. We introduce HerHealthEval, a controlled evaluation framework for multilingual understanding of women's-health communication. For each clinical case, HerHealthEval provides matched versions in English, French, and Modern Standard Arabic using six communicative forms: canonical, clinical, layperson, indirect or hedged, emotionally concerned, and deliberately under-specified. The first five express the same underlying concern and retain the same clinical information, whereas the under-specified form intentionally omits relevant details to test whether the model recognizes that clarification is needed. We evaluate a multilingual instruction model and QLoRA-adapted variants on concern classification, risk calibration, clarification behavior, parse compliance, and cross-form consistency. Results reveal that aggregate accuracy and consistency can conceal safety-relevant failures. A multilingual adaptation model reaches 0.994 under-triage in French and Arabic under language-asymmetric risk supervision. A controlled re-adaptation using source-derived, language-invariant risk labels reduces under-triage to 0.572 and 0.558, respectively. These findings show that robust multilingual healthcare evaluation requires explicit testing of register variation, uncertainty handling, and the provenance and invariance of adaptation labels.

Conversational LLM agents increasingly rely on Web search, yet the end-to-end lifecycle of agentic search remains poorly understood. We present the first study of Web search across four major conversational platforms (ChatGPT, Claude, Grok, and DeepSeek), combining real-world user interactions (invivo) with controlled experiments using the same platform's models by their APIs (invitro). We investigate the quality of agentic decisions to invoke Web search, their strategies to formulate queries, the potential domain preferences in the search results they receive, and the choices they make when transforming search results into grounded responses. We find that Web-search decisions vary substantially across platforms and models, while more frequent Web-search invocation does not necessarily yield better response quality. We further show that conversational agents employ different complex querying strategies and that platform specific search engines return search results from their preferred domains. Finally, although responses are largely grounded in search results, some claims rely on uncited search results, raising concerns about attribution and reliability. Our findings have important implications for the design of future AI agents and Web search tools optimized for conversational retrieval.

Operator learning for parametric partial differential equations (PDEs) typically builds global models over prescribed domains, requiring cross-condition data or costly physics-constrained training. Here we show that the output Jacobian of a neural solution model trained at one condition defines a reusable response space for cross-condition solution variations. We introduce Linearized Subspace Transfer (LST) to exploit this space and recover target solutions by minimizing the target PDE-system residual over response-space coordinates. Because any single response space has finite coverage, Active Transfer Modeling (ATM) uses post-transfer residuals as coverage indicators to selectively acquire response spaces from additional single-condition models. Across six systems, single-condition response spaces supported cross-condition transfer, with enrichment improving accuracy when added spaces expanded representation capacity. Relative to evaluated physics-informed operator baselines, ATM reduced error and offline construction cost, with orders-of-magnitude accuracy gains in representative cases and millisecond-to-second target adaptation. These results establish neural solvers as reusable local parametric models.

Generative AI writing assistants and the Large Language Models (LLMs) that power them are increasingly part of how voters gather information before elections. With growing evidence that they influence users' opinions, it is increasingly important to understand the views and positions of these tools. To better understand these views, we examine the stances supplied by six LLMs on a variety of Swedish-language writing tasks ahead of the 2026 Swedish parliamentary election. We cross 107 policy propositions with 77 writing templates and neutral, positive, and negative prompt framings, producing 24,717 prompts per model and 148,302 responses. To study these, we look at the models' default stance tendencies, compare how they respond to similar issues, and compare their responses with those of each of Sweden's eight parliamentary parties on the same issue. We find that Claude, DeepSeek, Gemini, and Mistral have similar profiles; ChatGPT more often supplies neutral or ambivalent text; and Grok differs most on topics such as migration, crime, and gender. When comparing the political parties, we find that the Social Democrats are closest to all six models. Still, after correcting for multiple comparisons, none of the within-model differences in party distances remains significant. Overall, we find that no model has a clear preference, nor a clear preference for a party, but that this depends on the specific issue or task the user asks about.

Cyber-Physical Systems (CPS) are commonly represented through multiple interconnected models. During development, CPS consistency requires that shared model elements remain compatible across these models. Uncertainty, for example, due to sensor noise or model abstraction, changes the admissible values of model elements and can introduce inconsistencies, i.e., situations in which models can no longer be jointly satisfied. While existing approaches can determine consistency for a given uncertainty configuration, they provide limited support for systematically exploring, analyzing, and explaining inconsistency across large uncertainty spaces. We address this challenge by reformulating inconsistency as an intervention response modeling problem. Using Saltelli sampling and multi-fidelity Monte Carlo estimation, we generate intervention-response datasets and train a surrogate model that directly predicts inconsistency from the propagated uncertainty geometry. Experiments on 48 scenarios and 10 CPS domains show that the surrogate matches Monte Carlo estimates while reducing evaluation time from milliseconds to microseconds, enabling orders-of-magnitude more response-surface evaluations within fixed computational budgets. Building on the learned response surfaces, we perform sensitivity analysis to identify dominant uncertainty drivers and introduce a gradient-based consistency recourse method to determine minimal uncertainty interventions that restore consistency. The results show that inconsistency under uncertainty can be effectively learned, analyzed, and repaired through response-surface modeling, providing a scalable foundation for uncertainty-aware consistency management in CPS development.

Railway engineers need simulation models that predict vehicle responses across operating scenarios that cannot be tested exhaustively. Agreement with representative measurements provides essential evidence, but calibration at a limited set of conditions does not guarantee accuracy elsewhere. We present a multifidelity railway-bogie response-correction method that treats multibody simulation histories as low-fidelity information and roller-rig measurements as high-fidelity evidence. This method combines an experiment-anchored fidelity assignment with physics-informed discrepancy learning for multichannel bogie-response histories. A time-delay neural network (TDNN) represents the condition-dependent simulation trend, and development-fitted amplitude alignment defines the low-fidelity baseline. A residual-correction network then models the reproducible response component not explained by this baseline and adds it to the baseline. An effective dynamic-balance equation constrains the learned discrepancy by representing differences in inertia, damping, stiffness, and external forcing between the simulated and physical systems. The training objective combines this constraint with residual matching, temporal smoothness, and a combined channel-2 acceleration loss selected using displacement-acceleration consistency evidence. For the evaluated reconstruction case, the corrected response gives a mean coefficient of determination of 0.8197, a mean normalized root-mean-square error (NRMSE) of 4.6055 %, and a mean normalized mean absolute error (NMAE) of 1.9297 %. These results provide initial evidence of accurate response prediction at the held-out 385 km/h condition.

In Retrieval-Augmented Generation (RAG) systems, a large number of retrieved chunks are concatenated to form the input context so that users can receive high-quality responses based on external knowledge. As a result, the input context length increases substantially, leading to a larger prefill workload and, in turn, a longer time to first token (TTFT). While previous works that reuse precomputed key-value (KV) caches effectively reduce TTFT for long-context inputs, it remains unclear whether response quality is preserved when the input context becomes very long. In this paper, we propose a combined approach that (i) fine-tunes the model while taking KV cache concatenation into account and (ii) selectively recomputes a subset of the KV caches. By applying both techniques, we demonstrate improved accuracy for long-context inputs. Experiments on the RULER benchmark show that, for a 124k-token input, our method improves the RULER score by 9.7 point over the baseline that recomputes KV caches only. Moreover, TTFT is reduced by 80% compared with full attention.

Benchmark scores describe what a checkpoint can do now, but they do not determine how it will respond to the next training episode. We measure this missing state by branching four short, standardized, target-independent micro-interventions from the same checkpoint and recording their effects in a common capability space. Together with current capability, these responses form L-State; its pulse block supports a flexible direct readout and a structure-preserving operator readout. Under smooth local dynamics, the operator construction admits an end-to-end cross-family bound with explicit source- and target-family coordinate heterogeneity. In three-family leave-one-family-out development, both pulse readouts reduce source-standardized MSE by 39.4% relative to capability alone, while separating the best response and direction estimates. On sealed GLM-4-9B, the direct and operator readouts reduce MSE by 71.8% and 78.3%, respectively, and the operator readout raises sign balanced accuracy from 0.366 to 0.754. On sealed Granite-3.1-8B, the direct readout reaches RMSE 0.544 and a development-fitted action-wise selector reaches 0.554, compared with 1.172 for capability alone. A five-family audit finds that the operator coordinate varies by action and family, and that modeling these deviations improves retrospective held-trajectory prediction. Target-independent interventions therefore expose training-response information that current capability misses, with direct and structured readouts covering complementary transfer regimes.

Understanding why individuals respond differently to psilocybin requires modeling the drug's transcriptional perturbation signature at the cell-type level. I present a Transformer-based delta expression encoder that learns to classify differential gene expression status - upregulated, downregulated, or neutral - from single-nucleus RNA-sequencing data, without supervision from pathway annotations or prior biological knowledge. The model is trained on pseudobulk profiles from 623 examples spanning 18 cell types, 2 drug conditions, and 6 timepoints derived from the Liao et al. 2025 dataset, and achieves 69.4% weighted classification accuracy. Three principal findings are reported, alongside one direct test of a published hypothesis that returned a result inconsistent with that hypothesis. First, per-cell-type classification accuracy ranges from 28.3% (L2/3 IT, a primary HTR2A-expressing psilocybin target) to 99.6% (endothelial cells), consistent with known psilocybin response biology. Second, psilocybin-induced transcriptional downregulation is significantly more stereotyped across individuals than upregulation (Mann-Whitney U=18615.0, p<0.0001), a novel finding with a cortical depth gradient across excitatory subtypes. Third, attention-guided gene co-regulation analysis recovers drug-specific modules without pathway supervision. Separately, a direct test of whether baseline HTR2A expression predicts drug-response separability across cell types found a significant negative correlation (Spearman r = -0.7088, p = 0.0021), the opposite of what a simple HTR2A-gating account would predict.

Large language models are increasingly used as sources of advice and information, including in high-stakes settings, yet little is known about how they respond to user disagreement. We study how a model manages its epistemic authority, referring here to its claim to knowledge, competence, or the right to advise, once a user challenges its answer. Building on Conversation Analysis, we introduce a taxonomy of six challenge types and a four-layer framework for analysing each response: whether the original claim is maintained or changed, where authority is located, how the disagreement is socially managed, and what kind of evidential support is offered. We construct a new dataset of 2,310 controlled challenge scenarios and 32,340 corresponding responses from 14 models, and analyse them using our framework with an LLM-as-judge pipeline, providing a vocabulary which future evaluation and benchmark design can build on. We find that models show conflicting behaviour: they validate users in 85% of responses but maintain their original claim in 65%. They explicitly apologise in 33% of responses, yet 59% of those apologies accompany maintenance of the original claim. They transfer authority most often in advice tasks, doing so in 28% of responses and reaching 57% in health advice and 49% in legal advice, compared with 6% in fact and 3% in explanation tasks. Abandonment of the original claim ranges from 0.8% for GPT-5.2 to 40% for DeepSeek 7B, while complete replacement of the original claim is rare overall at 1.5%.

Demographic synthetic survey panels are often validated by matching aggregate answers to published surveys. We test what that certificate establishes across six multiselect batteries from four survey organisations in three countries. The headline analysis is restricted to three instruments whose synthetic cohort and human target share the stated population frame; three other batteries remain sensitivity analyses. The response contract dominates measured fidelity. In the aligned instruments, committed sets leave 66 of 128 model-battery option slots empty in panels of up to 500 respondents, versus 0 of 128 under per-option probability elicitation. Across eight uncapped model-instrument comparisons, probabilities reduce option-marginal MAE by 4.53 to 7.30 points. The capped instrument reverses on two models until the vectors are projected onto its stated maximum. These are measurement effects: human targets are realised check-all responses, whereas the vectors are latent inclusion propensities. Published marginal agreement also fails to discriminate respondent simulation from direct population estimation. On nine aligned model-battery pairs, a no-persona population-prevalence query averages 6.27 MAE versus 12.39 for committed panels and wins all nine comparisons. Constraint-aware probability vectors average 5.34 and beat the query on four of nine, so the baseline challenges the validation criterion rather than proving direct estimation uniformly best. On three unpublished demographic cells, neither approach beats reciting the national distribution. Population-marginal agreement is therefore evidence about an elicitation contract and an estimand obtainable without simulated respondents, not evidence of individual simulation.

Conversational assistants can blend retrieval, action selection, escalation, and wording in a single model path, or separate those roles. We report a production migration of a customer-support assistant at a large accommodation marketplace (millions of conversations per month, 11 languages, 10-second P90). Dynamic Response (DR) replaces a single Qwen3-235B-A22B blended responder with a bounded ReAct orchestrator over typed tools plus a smaller generator that writes from a backend-validated context contract. Because the migration also changed prompts, alignment, and serving, we attribute each effect to its cause and claim as architecture effects only those measured on identical replayed turns: typed entity selection moves the reservation selector to a precision-first operating point (precision 8.3% to 89.1%, recall 75.2% to 67.3%), and typed action IDs with a membership check remove observed structured-action hallucination (2.14% to 0.0%). A low-ramp A/B test reproduces the replay escalation reductions: hard-escalation responses fall from 5.60% to 3.08% and soft-escalation responses from 9.56% to 2.49%, while production handoff volume holds roughly steady; self-solve is directional (+5.1 points, 95% CI [-2, +12]). Serving optimizations cut orchestrator P90 latency from 3.87s to 2.24s on a GPU footprint reduced by roughly one-third, and self-hosting reduces estimated annual model-serving cost by more than an order of magnitude.

Chatbots built on large language models (LLMs) are increasingly used as confidants. Tuned to satisfy users, they may answer with excessive empathy and affirmation that fosters dependence, and how the states and relationships of many users co-evolve under repeated consultation is hard to observe in real settings. We build a virtual classroom of 20 student agents who interact through rule-based chats, quarrels and consultations with friends and, when stressed, may instead consult a counselor AI (Gemini 2.5 Flash) under one of six style prompts: affirming, listening, solution-oriented, reality-redirecting, inciting and blaming. A second LLM call turns each exchange into updates of five state variables (stress, happiness, self-reliance, sociability, AI dependence) without seeing the prompt. We compare the seven conditions, including a no-AI control, over 15 and 50 days and under a lower consultation threshold, and test the robustness of the 50-day comparison with a pre-specified protocol: the same block in ten independent classrooms, repeated LLM realizations of one classroom with its event stream fixed, and evaluator updates scaled by 0.3 and 0.1. In every classroom the affirming and inciting prompts ended with lower self-reliance and higher AI dependence than the control, and the listening, reality-redirecting, inciting and blaming prompts with higher stress, lower happiness and more non-attendance; the solution-oriented prompt did not differ consistently from the control. The robust self-reliance and AI-dependence differences kept their signs at the 0.3 scale with highly similar rankings (Spearman 0.89, 0.93); the stress and happiness rankings did not, and the affirming prompt's lower stress reversed its sign. All quantities are simulation state variables, not effects on users. We specify the agent dynamics completely and discuss the limits of an LLM as generator of state updates.

Retrieval-augmented generation (RAG) can improve the specificity and grounding of large language model responses, but its effect is not uniformly beneficial in single-turn mental-health question answering, where user queries often combine emotional distress, treatment concerns, and safety-sensitive needs. We study when retrieval helps or hurts mental-health QA, and whether a lightweight selective retrieval policy can better control this trade-off. We operationalize retrieval need using three draft-conditioned utility dimensions: psychoeducational need, coping need, and response specificity, together with a rule-based safety trigger. Following psychotherapy-grounded RAG systems such as coTherapist, we construct a compact and controllable guideline corpus comprising coping-strategy, psychoeducational, and safety resources. We fine-tune an instruction-tuned generator on MentalChat16K using QLoRA and compare Closed-book, Always Retrieval, and Selective Retrieval settings on CounselBench-Eval and CounselBench-Adv. Experiments show that retrieval is not uniformly beneficial in this domain. Always Retrieval improves specificity but lowers overall quality and introduces additional safety-sensitive failures. Selective Retrieval preserves closed-book behavior for low-need cases while avoiding the additional degradation caused by unconditional retrieval, supporting the view that retrieval activation is a safety-sensitive control decision.

Massage therapy is helpful for the rehabilitation of various diseases, such as headaches caused by migraines and stress. Existing robotic systems have focused on massage therapy on the torso and limbs, but performing massage motions through suitable actuation on a person's head has been a challenge. In this paper, we present MiBOT, a head-worn massage robot that actuates two soft tactors to produce touch motions mimicking human massage. A key design principle behind MiBOT is its silent actuation, which we achieve through pneumatic artificial muscles in conjunction with a controller loop to respond to contact pressure. We evaluated the effectiveness of MiBOT in a controlled study and assessed subjects' blood pressure and heart rate levels while applying MiBOT. We found that our mechanical system generated positive and conclusive quantitative outcomes that are similar to the human-administered massage, decreasing participants' mean systolic and diastolic blood pressure by 2.8 mmHg and 1.7 mmHg, respectively, as well as calming their heart rate by 8-10% on average.

Multiple-choice question answering (MCQA) benchmarks in NLP use number-right scoring (accuracy), but in educational testing, the scoring scheme, the combination of the response mode models follow and the rule for grading responses, is a key design choice that dictates which abilities to reward. We examine how alternatives to number right change what MCQA measures with six education-inspired schemes that assess abilities beyond accuracy: distractor elimination, abstention, confidence calibration, and self-correction. On LLM benchmarks, these schemes: 1) shift rankings of 31 LLMs beyond rephrased number right prompts; 2) better predict the LLMs users prefer in LLM Arena; and 3) reveal distinct model capabilities, like that GPT-5 rarely abstains and readily self-corrects, while weaker open-weight models often abstain and hesitate to eliminate choices. Given the benefits of alternative scoring schemes, we discuss ways to extend them to tasks beyond MCQA.

Perturbation methods explain model decisions by measuring prediction changes under altered inputs, but response magnitude tells us only how much a model reacts, not what that reaction means. The same magnitude can support the final factual-counterfactual difference, oppose it, or arise strongly along the perturbation path yet vanish at the endpoint. We therefore track how the contrast develops as paired inputs are progressively revealed, using the final contrast to interpret the trajectory. We introduce DECAF (Decomposition of Evidence, Contradiction, And Fragility), which routes aligned, opposed, and endpoint-null responses into evidence E, contradiction C, and fragility F. The decomposition preserves ordinary magnitude exactly, Abs = E + C + F, and is unique under endpoint-relative axioms. Across controlled vision and tabular settings, the three components track independently measured behavior. In a 72-model ImageNet-9 audit, we compare cases with nearly identical response magnitude but different independently measured behaviors. The largest DECAF component agrees with an observed behavior in 96.4% of cases, compared with 35.0% for magnitude alone. Changing only the reveal path increases total response by nearly 80%, yet evidence barely changes while fragility grows by more than 4x. On FunnyBirds and ImageNet-1k, short forward-only DECAF trajectories outperform the tested general-purpose attribution baselines. On a 1B-scale DINOv2 model, a short trajectory matches a strong gradient-based baseline with 4.75x lower wall time and 2.36x lower peak memory.

Hybrid-thinking multimodal large language models (MLLMs) allow a single model to alternate between deliberative thinking and latency-efficient non-thinking inference. Although these modes differ in reasoning budget, their delivered responses should satisfy the same user-facing standard. Correctness alone may not characterize this response quality; we therefore evaluate task accuracy and response-pattern failures as complementary outcomes. We study this gap through \textbf{response-pattern alignment}: whether thinking and non-thinking interfaces preserve acceptable final-response behavior. We introduce \textbf{PatternEval}, a failure-enriched diagnostic benchmark comprising 2,415 multimodal prompts spanning visual perception and grounding, structured image understanding, and multimodal knowledge reasoning. PatternEval tests four recurrent failures: chain-of-thought leakage, response repetition, logical contradiction, and performative reasoning. Response-pattern failures are widespread across models from different providers, with non-thinking inference exhibiting substantially higher failure rates and thereby creating systematic misalignment between thinking and non-thinking interfaces. Motivated by this diagnosis, we develop \textbf{PatternRM}, a response-level reward model, and \textbf{PatternRL}, which introduces pattern-specific penalties during reinforcement learning. Experiments on Qwen3-VL-4B and Qwen3-VL-8B show that incorporating pattern-specific penalties into reinforcement learning can mitigate cross-mode misalignment while incurring a marginal task performance trade-off. Together, PatternEval and PatternRL provide an evaluation-and-training framework for aligning user-visible response patterns across hybrid-thinking interfaces.

Is it possible to find an objective classification criterion that organizes the complexity of gust-induced loads across many flight conditions? And one that remains as interpretable as a labelling based on coarse parameters, such as the flight attitude? Our approach encodes a large number of experimental observations through a machine-learned representation and applies a summarization procedure to select a minimal subset of highly significant exemplars. The exemplars provide a similarity-based objective classification criterion of all the observations, they can be more conveniently inspected by experts and can become subject of more refined experiments. We demonstrate the approach on a database of 3480 pressure-load measurements induced by random gusts on a flying-wing model across six flight attitudes. We find nine fundamental response types that recur across multiple attitudes; analysis of a type's transient response enables physical intuition into the underlying fluid mechanics.

Large Vision-Language Models (LVLMs) integrate visual perception with language generation, enabling responses that span image understanding and complex reasoning. However, LVLMs do not just inherit the text-level hallucinations; they also hallucinate against the image, producing fluent responses ungrounded in what they see. This makes LVLM response scoring inherently harder, and our diagnostics show that existing confidence-based metrics adopted from LLMs are insufficient for LVLMs. Specifically, removing the input image barely changes confidence-based selection, suggesting that output-space confidence primarily captures textual plausibility rather than agreement with the image. To address this gap, we propose LookBack, a training-free LVLM response scoring method that augments token likelihood with visual lookback score, a lightweight measure of how strongly each response token refers to image tokens. Across four benchmarks and three models, LookBack consistently improves Best-of-$N$ selection over existing baselines with negligible additional overhead.

Drug response prediction (DRP) models are an active area of research in pharmacogenomics, with growing potential to accelerate the identification of effective anticancer drugs. However, their predictive performance is often constrained by limited dataset scale and insufficient coverages of cancer and chemical spaces. In addition, inconsistent benchmarking practices hinder reliable comparison across models. Standardized frameworks, such as the Innovative Methodologies and New Data for Predictive Oncology Model Evaluation (IMPROVE) project, provide unified data schemas and evaluation protocols for consistent benchmarking, but improving model generalizability requires larger and more diverse training data. In this work, we substantially expand the IMPROVE benchmark through large-scale integration of pharmacogenomic data, primarily from PharmacoDB, together with additional smaller data sources. The expanded resource includes millions of drug response measurements, broader multi-omics coverage, and a major increase in chemical diversity, adding more than 50,000 compounds. To evaluate the impact of the new dataset compared to the original IMPROVE benchmark dataset, we trained DRP models using the two datasets and assess their prediction performance using a common test set and several evaluation strategies, including drug-blind, cancer-blind, and disjoint data splits. While cancer-blind performance remained comparable to the original benchmark, models trained on the expanded dataset showed consistent improvements in drug-blind and disjoint settings, indicating enhanced generalization to previously unseen compounds. These results position the expanded dataset as a community resource that provides a richer foundation for developing DRP models intended to aid in the discovery of novel anticancer drugs.

Human-AI research often evaluates individual capabilities, combined performance, or final outputs, but these approaches do not preserve how one party's response becomes part of the conditions under which the other party's next contribution is formed. This article introduces socioduality, a sequential, reciprocal, and history-carrying relational process between two distinguishable parties in which a response from one party becomes part of the observable conditions under which the other party's subsequent contribution, judgement, decision, or action is formed. Specified for human-AI dyads, the construct uses nested units: moves, confirmed sociodual episodes, linked pathways, and the broader interaction container. A minimum episode A1-B1-A2 requires evidence of response contingency and return contingency; candidate episodes are classified as confirmed, non-sociodual, or indeterminate before secondary coding of response orientation and substantive contribution re-formation. Three propositions address history-conditioned formation, pathway divergence, and robustness differences among endpoint-equivalent pathways. A frozen operational protocol was calibrated on three previously unseen natural human-AI records through two separately executed model-based evaluator series. Move and candidate reconstruction converged exactly in two cases and differed by one local multimodal unitisation decision in the third; remaining disagreement was concentrated at return-contingency boundaries. Socioduality therefore provides a bounded and empirically tractable process construct for analysing how human and AI contributions are formed through interaction while preserving pathway information that endpoint-centred analysis cannot recover.

Online credit card discussions provide a natural setting for studying how consumers communicate about financial products. Simulating these discussions requires more than just generating individual comments, the generated threads should also match how real users express themselves and interact with others. We introduce CARD, a framework for generating realistic credit card discussion threads. Given a credit card post and its matched real thread, CARD uses non-verbatim guidance on reply structure, comment function, stance, tone, and conversational variation. A planner organizes these controls, a writer generates the discussion, and a calibration loop updates comments' populations that contribute to differences between the generated and real thread distributions. We evaluate CARD on real Reddit credit card discussions using lexical, semantic, behavioral, and structural metrics. CARD matches the distributions of real credit card discussions better than simulation baselines across multiple LLMs and also demonstrates smaller effect sizes and distribution distances across metrics. These results show that structured planning and targeted revision can generate the realism of simulated credit card discussions.

A failed retrieval-augmented generation (RAG) answer can be consistent with several unseen responses to evidence repair. We introduce Pair-ID, an offline audit that holds one query, retrieval state, and reader constant, then crosses two operations, adding missing support and deleting verified nonsupport, to measure a same-failure counterfactual response vector. A complete funnel over 19,981 benchmark queries identifies 11,105 eligible Qwen failures, from which a prospectively fixed SHA-256 ordering selects 1,200 before generating any sampled response. Among 1,190 regenerated-valid failures, support addition repairs 197/600 JOINT cases (0.328, 95% CI [0.292, 0.367]), and deletion repairs 162/1,190 cases (0.136, 95% CI [0.117, 0.155]); length- and position-matched shams retain semantic contrasts of 0.223 and 0.101. The original view carries partial predictive signal for individual response cells (macro AUROC 0.678; Brier 0.152 versus 0.160 for a marginal baseline), but exact-vector accuracy, 0.637, does not exceed the 0.646 majority-vector baseline, and vector macro-F1 is 0.170. Across four readers, both marginal sensitivities recur, while pooled exact-vector agreement is 0.675-0.765 and JOINT-only agreement falls to 0.538-0.691. These results show that evidence sensitivity occurs at meaningful rates in the hash-selected eligible-failure sample, is only partially predictable from the observed failure, and is conditional on the reader. The evidence supports a frame-scoped offline response audit, not an information-theoretic impossibility result, reader-independent taxonomy, or runtime repair policy.

Musical interaction with AI is often organised as a response loop: a human performs, the system interprets that action, and the system answers, accompanies, or schedules a musical event. Unconducted vocal ensembles pose a different problem. Singers act simultaneously and continuously affect one another; neither timing nor pitch is fixed by a conductor, metronome, accompaniment, score, or tuning source. Collective organisation emerges from many-to-many reciprocal adjustment. This paper frames such ensembles as coupled dynamic systems and proposes a research architecture for vocal agents that enter, rather than merely track, their collective states. Some target repertoires are metrical, while others exhibit non-isochronous temporal contours that cannot be reduced to a beat grid; we treat the latter as a hard case for a general framework. The architecture connects multichannel capture in the field to dialect- and singing-aware representation, collective-state inference, vocal generation, and in-situ evaluation. The resulting agenda asks not only whether an artificial singer can synchronise, but how its presence reorganises human coordination, leadership, style, and musical transmission.

Audio-language models (ALMs) are increasingly used as judges for speech-to-speech systems, but a judge that receives audio may not actually use paralinguistic evidence. We introduce counterfactual audits for paralinguistic response evaluation. Each audit item holds the transcript fixed while varying affect, prosody, or the timing of an affective shift, forcing a valid judge to track the audio cue rather than lexical content or response style. We evaluate ALM judges using a native one-context judgment protocol and a contrastive recoverability control, then further decompose each item into its constituent perception and response-mapping skills. This yields useful diagnostic states that identify different sources of judge failures. Across Gemini, GPT, and open audio models, we find that contrastive success often overstates native judge reliability, and that similar aggregate accuracies can hide different failure modes. These results suggest that ALM judges should not be evaluated by accuracy alone, instead requiring thorough behavioral audits before deployment.

Language models are often evaluated as though capabilities demonstrated in English remain equally available when the same content is presented in other languages. Traditional multilingual benchmarks rarely isolate language while holding content, question, reference answer, model, and evaluation unit constant. We define the Cross-Lingual Comprehension Gap (CLCG) as the reduction in response quality when the same content and question are presented in a target language rather than in English. Using ParallelQA-18, a professionally human-translated parallel corpus, we evaluate five models from five laboratories on a stratified sample of 150 articles across 18 languages (English reference; Portuguese high-resource baseline; 16 targets spanning Joshi et al. 2020 classes 0-4). A within-item design varies only passage language. The primary estimator contrasts English versus pooled target-language Token-F1 micro-means on higher-complexity open-ended questions, with article-cluster bootstrap intervals. The primary pooled CLCG is 0.078 (95% CI 0.072-0.084), about a 17% reduction relative to the English score; the equal-language macro summary is 0.077. Net of Portuguese, the macro gap is 0.016 (95% CI 0.013-0.020). Language-level CLCG is negatively associated with Joshi resource class (rho = -0.594, p = 0.015, n = 16). In blinded paired human evaluations, higher-resource responses are preferred in 61.6% of decisive judgments (estimated preference probability 0.655, 95% CI 0.558-0.741). Capabilities shown in English should not be assumed to transfer equally to other languages; English-centered evaluations may overestimate quality for users of low-resource languages.

Large language models (LLMs) can generate fluent and convincing text at scale, creating growing risks for misinformation dissemination, educational misuse, and platform governance. These concerns make robust detection of machine-generated text increasingly necessary. Recent zero-shot detectors mainly exploit probability-based statistical discrepancies, but they do not explicitly account for the training process of LLMs, which leaves a distinct generation mechanism insufficiently modeled and limits detection robustness. To address this issue, we propose EchoPrompt, a training-free detector based on latent prompt restoration. Our key intuition is that machine-generated text is typically produced conditioned on an upstream prompt, and this hidden dependency can be partially reactivated by prepending a unified generic prefix. Specifically, EchoPrompt restores a generic assistant-response context, measures the induced likelihood gain with an instruction-tuned model, calibrates it against the corresponding base model, and aggregates the resulting differences into a score that quantifies latent prompt dependency. Extensive experiments show that EchoPrompt achieves state-of-the-art performance among zero-shot detectors while maintaining strong robustness across challenging evaluation settings.

Language models differ in how safely they behave and these differences are measured by safety benchmarks. But aggregated benchmark scores are hard to trust and interpret, because benchmarks duplicate one another, correlate heavily, and models may sandbag when they detect evaluation. To address these issues, we draw on Item Response Theory (IRT), a statistical toolkit for measuring these latents from performance on items with inferred psychometric properties. We fit IRT models to eight safety benchmarks across 192 language models, the largest psychometric analysis of LLM safety evaluations to date, and contribute three results. First, we find that three interpretable factors of refusal strictness, truthfulness, and contextual harm explain most of the variance between models across benchmarks. Second, psychometrically selected items recover full benchmark scores with lower error than random subsets of the same size, and roughly ten adaptively chosen items suffice for several individual benchmarks, cutting evaluation cost by 97-99%. Third, IRT supports audits of individual models, showing that it can be used to detect naive sandbagging and changes of model behind APIs. Overall, we show IRT is a ready-made toolkit for reading, reducing, and auditing safety benchmarks, which we recommend frontier labs and evaluators adopt.

Black-box language-model reliability is commonly pursued by sampling, prompting, voting, verifying, or iteratively revising individual answers. We ask a prior question: \emph{what determines whether a collection of black-box responses is recoverable at all?} We represent responses to typed transformations of a query as a \emph{relational response field} (RRF). Edge transports encode how valid responses must change under paraphrase, scaling, decomposition, refactoring, or other task symmetries; anchors encode independently trusted evidence such as execution or a verifier. For relation operator $D$, anchor operator $A$, and at most $k$ corrupted response nodes, we identify $γ_k(D,A)$ as the intrinsic difficulty of black-box response recovery. It is positive exactly when every $k$-node corruption is identifiable; it gives a deterministic stability bound proportional to $1/γ_k$; and a matching two-point minimax lower bound shows that no estimator can improve this dependence. Thus consistency is not truth: relation-only methods are blind to null directions, including shared hallucinations. We derive sparse field-repair algorithms while separating information-theoretic identifiability from the stronger null-space conditions required by convex optimization. Controlled theorem tests and black-box mathematics/code experiments evaluate four theory-fixed consequences: consistency--truth separation, anchor phase transitions, redundancy saturation, and cross-model, cross-task prediction of repair difficulty. The results support $γ_k(D,A)$ as a measurable property of a response-recovery instance, rather than a score attached to one repair heuristic.

Most multiple-choice question (MCQ) benchmarks evaluate Large Language Models (LLMs) only by whether they select the correct answers. This binary scoring treats all incorrect responses alike, even though an LLM's preferences among incorrect options may contain systematic and useful information about its behavior and ability. We introduce the LLM Nominal Response Model (LLM-NRM), an option-aware psychometric framework that models the full distribution over answer choices to jointly estimate LLM ability and option-level item characteristics, while separating model-specific response calibration sharpness, positional preference, and difficulty-dependent fallback behavior. Across 189 LLMs and 31,554 items from 14 benchmarks, LLM-NRM predicts held-out LLM-item interactions more accurately than binary Item Response models and conventional nominal-response baselines, and its ability estimates achieve the strongest Spearman correlation of 0.920 with the external human-preference Arena.ai Elo leaderboard. Distractor identity contributes +101% additional Fisher Information per item beyond correctness, and incorrect responses alone recover full-information ability estimates with Spearman 0.943. The learned item parameters also enable efficient benchmarking, where 41 selected items preserve the full-bank ranking with Kendall's correlation 0.85, corresponding to a 770 times reduction. In conclusion, we show that incorrect answers carry distinct and useful measurement information rather than representing equivalent mistakes.

Predicting transcriptomic responses to small-molecule perturbations across cell lines is central to drug discovery, but exhaustive profiling of drug-cell combinations is infeasible. We frame molecular perturbation prediction as retrieve-and-aggregate: approximate an unmeasured drug's response in a cell line by aggregating measured responses of a small set of biologically related compounds. We propose LLM-Guided Retrieval (LGR), where a large language model (LLM) ranks candidate neighbor drugs (restricted to those profiled in the target cell line); after which a fixed mean aggregator combines their observed expression deltas to form the prediction. We evaluate on the Tahoe-100M single-cell perturbation atlas under unseen-drug, unseen-cell-line, and open-world regimes. LGR consistently improves over drug mean, ChemCPA, and chemistry-based kNN baselines, with the strongest gains for unseen cell-line generalization, where it achieves higher correlation and lower error than mean baselines. Across settings, LGR improves directional (sign) accuracy of gene regulation, indicating better recovery of biologically meaningful perturbation effects even when magnitude-based metrics are similar. These results suggest that retrieval quality, rather than predictor complexity, is a key driver of zero-shot molecular perturbation prediction, and that LLMs can provide a useful biological prior when used as constrained retrieval modules.

Memory-augmented agents can know that a user's stored state is outdated and still plan around the old value. The STALE benchmark calls this the implicit policy adaptation (IPA) gap. We identify one structural contributor: draft-anchored verification checks what a response says, and in an open-ended response the stale dependency is usually unsaid. StateAuditor therefore audits in the opposite direction, from stored state to draft. An LLM proposes candidate old-to-new transitions from timestamped evidence; deterministic code pins each quotation to a single entry, checks that the new evidence really is newer, and lets only these verified transitions trigger repair. What is verified is provenance and chronology - not semantic supersession. On STALE's full protocol (400 scenarios, 50-session histories, one independent response per query), strict single-query VTA scores .736 against .686 for our locked predecessor under the same judge: a +5.0-point paired gain (95% CI [+2.9, +7.2]) coming almost entirely from IPA and premise resistance (PR). The benchmark's own judge, from a third model family, reproduces the gain (.738 vs. .680). On an independent cross-family preference-evolution benchmark (HorizonBench), the full draft-audit-repair pipeline over a gold-derived structured store raises current-preference accuracy (user-clustered p<.01), though a matched control shows most of this external gain is the draft-side audit itself; a harder authored lifecycle set gives no gain, bounding the claim while false invalidation stays controlled. On STALE, by contrast, a matched control (same evidence, adapter, and call budget) scores only .692 (+0.6 over the predecessor, n.s.), attributing the STALE gain to the transition machinery rather than added context or calls. We make no claim about general-purpose agent memory.

An agent playing a Nash-equilibrium strategy in a two-player zero-sum imperfect-information game secures the game value but forfeits the additional value offered by a flawed opponent. Diffuse deviations pose a particular challenge: binary release rules may gather too little evidence to act, while a full best response to an incomplete opponent model can be highly exploitable. We introduce \emph{budget-constrained confidence-scheduled restricted responses} (CS-RNR), the first opponent-exploitation method whose safety guarantee is a certificate the agent computes on the strategy it actually deploys, so that every exploit it commits to is one it has audited itself. The method tracks pooled action frequencies with anytime-valid confidence sequences and treats a frequency as exploitable only once its interval separates from an equilibrium reference. The confirmed deviations define a conservative opponent model, which a restricted-response solve turns into candidate counter-strategies over a grid of pin levels. Before deployment, each complete candidate is evaluated by a full-tree best response. The resulting certificate is compared with a user-specified budget and committed atomically with the strategy. Because this check is performed on the played strategy, model quality determines the exploitation achieved while the certificate controls reference-relative expected loss. In Leduc hold'em, CS-RNR obtains $6.2\times$ the steady-state gain of a money-verified binary gate while keeping every deployed strategy within budget. A trajectory mixture using the same estimator reaches $13.6\times$ the budget. Across Leduc, Liar's Dice, and 5-rank Leduc, all $36{,}000$ audited hands satisfy the reported certificate tolerance.

Supervised fine-tuning (SFT) can equip large language models (LLMs) with domain knowledge for high-performance computing (HPC) tasks such as data race detection and benchmark question answering. However, knowledge alone does not guarantee task-appropriate behavior: the same SFT model that correctly classifies 88.65% of C/C++ data race samples produces verbose, imprecise answers to factual queries, with 65.9% of MLPerf responses exceeding 40 characters. Reinforcement learning (RL) post-training addresses this gap by optimizing for task-specific rewards rather than token-level imitation. Yet HPC tasks exhibit extreme heterogeneity, with binary classification, factual QA, and semantic generation differing by 58x in answer length, spanning three distinct reward distributions, and showing widely varying SFT accuracy. This makes uniform-weight RL methods such as GRPO suboptimal. We propose HARGO, Heterogeneity-Aware Reward-Guided Optimization, which introduces per-response importance weighting via confidence-modulated advantage: computing a discrimination signal from group-level reward contrast and a confidence signal from reference model log-probabilities, then modulating the advantage before computing per-response weights, without requiring task-type labels. Across four HPC tasks and nine methods, HARGO achieves the best performance on all three primary metrics: WinRate 54.62%, Data Race F1 91.30%, and PLP Similarity 0.8558. Ablation confirms complementary contributions from both signals. HARGO establishes the best overall alignment quality among compared methods for heterogeneous HPC tasks.

Single-cell perturbation atlases rarely measure every intervention in every cellular context: a query perturbation is often observed in one or more source contexts but missing in the recipient context where its effect is needed. Ignoring those measured responses discards query-specific experimental evidence, whereas copying or weakly calibrating them across contexts risks transferring the wrong signal. We propose PerturbMap, which predicts a missing recipient-context effect by combining a recipient-local low-rank base with accepted proposals that transport the same perturbation's measured source responses through source-to-recipient ridge experts fit on paired training perturbations, with proposal weights determined by route reliability estimated on validation anchors. On the Perturb-CITE-seq melanoma cohort, PerturbMap improves full-effect MSE by 4.1% over a recipient-local low-rank base and achieves lower MSE than FedAvg, zero-response, raw-copy, calibrated-copy, and identity-shuffled affine controls. It remains within $2.82\times10^{-6}$ MSE of our centralized token-matched pooled reference, which uses a stronger training interface. A condition-mean specificity diagnostic shows the same direction: same-recipient top-10 counterpart retrieval by cosine increases from 74.5% for the low-rank base to 80.5% for PerturbMap.

As search increasingly shifts toward LLM-driven answer engines, advertising is becoming embedded within the generated response itself and should therefore be evaluated for both user utility and commercial value. The key challenge is click-through intent: behavioural logs are unavailable, human annotation resists calibration, and frontier LLM judges conflate intent with linguistic fluency. These gaps compound, as principled pricing presupposes a continuous intent signal, while generating such a signal presupposes supervision that is currently unavailable. We construct the missing supervision through a psychologically grounded agent simulation framework, and distil it into a parameter-efficient evaluator that predicts click-through intent, together with the three companion dimensions of ad quality, as smooth, differentiable estimates. Validated through sign-certain behavioural perturbations, the evaluator surpasses frontier zero-shot judges on relevance sensitivity (79% versus 60-67%), tracks graded content degradation, generalises without error to 103 fictional products, and agrees with human preference in 86% of pairwise judgements across five annotators, with agreement rising in the evaluator's confidence. Upon its estimates we build the pricing layer directly, deriving the unique payment rule under which truthful bidding is optimal, demonstrating it on a best-of-k allocation, and extending the mechanism to non-monotone allocations. The same differentiable signal stands ready as a training objective for ad generation.

Classical spoken dialogue systems often separated dialogue management from response realization: a policy selected the next dialogue action, and a generation component expressed that action. As dialogue systems shift toward LLMs, this decomposition has largely disappeared into the model's hidden representations. We ask whether an LLM-internal analogue of state estimation and action control can be recovered for conversational moves such as acknowledging, checking, querying, explaining, and replying. We formulate move control as two coupled problems: selection, predicting the appropriate next move from the dialogue context, and realization, causally producing a chosen move at generation time. We introduce Latent-IM, an internal dialogue-management framework that provides a general interface for choosing and deploying conversational moves under different objectives. Here, we use this control to reproduce human move choices, improving average end-to-end move accuracy by 12.5 points over the unsteered backbone while performing comparably to fine-tuning.

Test-time reinforcement learning (TTRL) enables language models to self-evolve at inference time without labeled feedback. Existing methods rely on answer voting and therefore do not extend naturally to open-ended generation, where valid responses cannot be mapped to a shared canonical answer. Without external reward models or stronger judges, adaptation must instead construct reliable rewards from the model's own outputs. We introduce SERPO (Self-Evolving Rubric Policy Optimization), which replaces answer voting with a closed loop that co-evolves response evidence, query-specific rubrics, and policy parameters. Good-Normal-Bad (G-N-B) response evolution organizes maximally separated rollouts into ordered archives; rubric evolution retains criteria that discriminate these archives; probabilistic criterion scoring converts verdict-token likelihoods into reward signals; and policy evolution optimizes the actor with the resulting signals. New actor rollouts then refresh both the archives and rubrics, closing the three-way evolution loop. Across two model configurations, two in-domain benchmarks, and four OOD benchmarks, SERPO improves HealthBench and ResearchQA by up to 20.63 and 20.31 points over the corresponding base models, raises the six-benchmark macro-average by up to 8.06 points, and supports OOD transfer and continued cross-benchmark evolution.

Large language models (LLMs) are increasingly used as synthetic users, stand-ins for human respondents whose simulated answers feed product, policy, and market decisions. We ask when this substitution is valid and when it fails, and package the answer as an evaluation framework for intelligent synthetic-user systems. A single protocol, run across four models spanning two families and an 8B-to-frontier capability range, is applied to two independent domains of real human-response data: U.S. general social attitudes (General Social Survey) and cross-cultural values (World Values Survey). Every model is benchmarked against a suite of non-LLM baselines fit on held-out human data. Under demographic prompting and the survey-simulation protocols we test, two failures replicate across both domains, all four models, and both families. First, at the individual level no LLM beats even the strongest baseline; on cross-cultural values every model falls well below it, and the gap survives distance-aware and proper scoring. Second, models systematically over-determine demographics, treating identity as far more predictive of attitudes than it is among real people, a distortion present for nearly every question-group combination and robust to a coding-invariant measure. Neither failure is remedied by a larger, more capable model. A decision-impact analysis shows why this matters in practice: on a segment-targeting task the models inflate between-segment gaps two to fourfold, would direct a team to the wrong segment in half of U.S. and most cross-cultural cases, and manufacture segment splits that do not exist in real people. We make the cross-domain benchmark and the evaluation framework available on request, so that teams can determine in advance when synthetic-user evidence is safe for decision support and when it is not.

Content-moderation classifiers are usually evaluated in isolation, but deployment requires choosing where to intervene and what follows a flag. We evaluate these choices using two end-to-end customer-outcome metrics rather than component accuracy: Usefulness, the fraction of turns with a shown, non-harmful, relevant response, and Harmful Exposure, the fraction with a shown harmful response. Latency and error rates are diagnostics. We compare Input only, Response only, and Input + response hard blocking on a human-labelled product benchmark and public ToxicChat evaluation. At the evaluated operating points, Response only achieves the highest filter-only Usefulness in both settings, while Input + response achieves lower Harmful Exposure. Replacing Response only blocking with Response + rewrite recovers most blocked traffic and yields the same observed Harmful Exposure count as Response only blocking for the selected configuration; this equality is not an equivalence result. Probe routing substantially reduces conditional route-and-generation time relative to LLM routing at comparable measured outcomes. A focused output review shows how rewrites balance filter passage with usefulness by generalizing triggering language while retaining benign intent and safe redirection; some sensitive-domain outputs nevertheless omit potentially safety-relevant support information. These results support comparing moderation configurations under deployment-specific safety and latency constraints rather than applying a universal placement rule. Code and public artifacts are available at https://github.com/microsoft/mod-frontier

Multimodal drug discovery enables drug representation learning beyond chemical structure by incorporating cellular responses such as gene expression and cell morphology. However, direct fusion and instance-level contrastive alignment may mix mechanism-related signals with modality-specific noise and incorrectly separate structurally dissimilar but biologically related compounds. This limitation can obscure transferable mechanism patterns required for predicting the properties of unseen compounds. We introduce PMRD, a pharmacological response domain-guided framework for multimodal zero-shot drug property prediction. PMRD separates mechanism-consistent factors from modality-specific information and constructs a consensus response domain across three modalities. Mechanism candidate augmentation identifies locally stable factors, while retrieval-geometry attribution dynamically reweights the alignment and augmentation objectives according to whether their updates preserve inter-drug discriminability.This feedback suppresses training signals that conflict with mechanism-discriminative retrieval. PMRD further combines complementary representations through reliability-aware multiview retrieval. Experiments on public datasets show improved zero-shot property prediction and more biologically coherent drug neighborhoods. Hard-negative analysis further indicates fewer conflicts between structurally dissimilar but response-related compounds. These results support PMRD as an effective framework for mechanism-aware multimodal drug representation learning.\footnote{The code will be released upon publication.}

Conformalized selection has been widely applied to select high-quality candidates from large datasets with rigorous uncertainty quantification, such as reliable labeling, drug discovery, and the alignment of large language models. Nevertheless, existing methods assume clean responses on calibration data, an assumption that rarely holds in practice. In this paper, we formulate the above tasks as selecting candidates with true predicted labels or with responses exceeding certain values. We demonstrate that existing conformal selection methods fail to control the false discovery rate (FDR) or suffer from severe power loss under contaminated calibration data. To that end, we propose Robust Conformalized Selection (RCS), a unified framework for selective classification with valid FDR control under general label contamination. The key insight of RCS lies in a novel statistical reduction: by separately conditioning on different classes, we translate the intractable label noise into a localized covariate shift problem, which then enables a covariate-adjusted empirical-Bayes-type estimate of the number of false selections. Statistical properties such as the asymptotic FDR control, power optimality, and robustness of RCS are established. We further develop an instantiation of RCS under randomized response model, and also apply RCS to the task of selecting candidates with large response values. Extensive experiments on both simulated and real-world datasets demonstrate the effectiveness of RCS.

Recent large reasoning models often develop long chain-of-thought responses during reinforcement learning (RL), resulting in high inference latency and deployment cost. Existing methods for response length control typically rely on explicit length penalties or additional control modules, which require careful tuning and may compromise reasoning quality. We propose Quadrant-weighted Sampling for Length-aware Policy Optimization (QLPO), a simple resampling-based variant of GRPO that introduces implicit length control without modifying the reward function. QLPO first over-generates candidate responses and then resamples the training group by preserving the empirical correct/incorrect ratio while favoring short correct responses and long incorrect responses. This reshapes the training distribution and implicitly encourages shorter model outputs. Across models ranging from 1.5B to 32B parameters, including both base models and strong reasoning models, QLPO consistently improves the accuracy-length trade-off. It reduces response length by 30% to 70% while preserving reasoning performance. These results suggest that structured resampling provides an effective and robust approach to efficient reasoning.

Large language models (LLMs) have shifted human--computer interaction from traditional'' interface journeys toward more conversational exchanges. Researchers studying HCI and UI use moderated usability sessions, interviews, surveys, transcript analysis, and static prototypes. However, static prototypes provide limited opportunities to study interaction with live AI systems or systematically control how an LLM behaves across participants and scenarios. Conversation transcripts reveal little about how users formulate, revise, and hesitate over prompts before submission. We designed the Human--AI Research Platform (HARP) for researchers, designers, and anyone who has ever wondered, What if AI did this?' HARP places participants in controlled mock scenarios with live, configurable AI agents. Researchers can control agent prompts, model parameters, response characteristics, and experimental conditions; trigger surveys at predefined moments; and record prompt composition time, response latency, deletions, and keystroke pauses. Planned capabilities include voice, facial expression, gesture, and, where legally and ethically appropriate, emotion analysis. We illustrate HARP through a study examining how technical specificity and response length affect retention of LLM output. By pairing controllable live agents with behavioral and self-report measures, HARP enables systematic testing of how AI design choices affect users.

LLM-based game agents often perform poorly on more complex tasks. This work examines whether these failures are linked to limited spatial reasoning and evaluates whether causal prompt augmentation and multi-step planning can improve win-rates while managing response latency. Using the open-source Qwen3 model family, we conduct experiments across varying model scales, reasoning modes, and planning horizons. We further introduce a focused GVGAI benchmark consisting of three custom games with five difficulty levels to isolate spatial navigation. The evaluation follows two paradigms: an initial ``positioning experiment'' to test an agent's ability to find its exact coordinates, and a study of game-play success. Our results show that while larger models with an enabled thinking mode identify their positions more accurately, overall performance in coordinate matching remains limited for smaller models. Win rates decrease as game levels and layout complexity increase, validating the benchmark's difficulty scaling. Integrating causal context into the prompts tends to improve the agents' success rates, particularly for bigger models. While enabling thinking mode and longer planning horizons significantly improve performance, multi-step planning further reduces mean per-step response times, offering a practical trade-off between reasoning depth and execution speed.

Users frequently express their beliefs to large language models (LLMs). In some situations, the LLM should accept these contextual beliefs as true. In others, they should stick to their prior knowledge. Notably, users' expressions of belief (EoBs) can take linguistically diverse forms - using presuppositions, evidential and certainty markers, or varied tones - each of which may have a different persuasiveness over the LLMs. We introduce a typology to systematically evaluate how different EoBs affect whether models follow context versus prior knowledge. The typology is grounded in four linguistically motivated dimensions: form, evidentiality, epistemic stance, and tone, spanning 17 fine-grained types. By pairing these EoBs with world knowledge facts, we generate controlled EoB-query pairs that isolate the effect of linguistic variation. Using this benchmark, we evaluate 16 LLMs that differ in architecture (Llama3, Qwen3, Gemma3), scale (1B-30B parameters), and training stages (base vs instruct). We identify meaningful variations in response behavior across these axes, e.g., that bigger models and instruction models tend to be less context-following than smaller models and base models. We further identify specific EoBs that statistically significantly persuade LMs more consistently than others. Our work reveals systematic patterns in how linguistic framing affects LLM context integration, with implications for prompt engineering and model robustness.

Proactive agents are expected to anticipate user needs and provide autonomous assistance by perceiving environmental context without explicit instructions. A fundamental capability of such agents is to identify and track users' upcoming events, enabling continuous and event-specific assistance. For example, by recording the time and location of a planned hike, an agent can deliver weather reminders in advance or provide navigation support before departure. However, existing works on proactive agents largely overlook event-centric assistance, and the open-ended nature of proactive assistance poses challenges for reliable evaluation. To bridge these gaps, we introduce ProEvent, the first event-centric benchmark designed to assess an agent's ability to proactively maintain a user's timetable based on ongoing instant messaging chats. ProEvent provides synthesized yet realistic chats that consider the dynamic interaction among users, concurrent chat threads, and noise in the real world, and evaluates proactive agents on response timing, single-step response correctness, and multi-step response correctness. Experiments on eight LLMs and pipelines reveal that current agents frequently overact and struggle with event cancellation. Notably, even GPT-5.1 only reacts correctly in 26.7% of scenarios. Further qualitative analysis reveals fundamental limitations of current LLMs as proactive agents, particularly in detecting implicit events and reasoning from the user's first-person perspective.

Multidimensional graded response models (MGRMs) are widely used for analyzing ordinal questionnaire data in psychological and educational assessments. A central challenge in applying these models is determining the number of latent dimensions. Conventional approaches usually fit multiple fixed-dimensional models and select among them using post-hoc criteria such as AIC, BIC, or cross-validation, which can be computationally demanding and ignore uncertainty in dimensionality during estimation. We develop an adaptive Bayesian dimension selection framework for probit MGRMs. Building on the cumulative shrinkage process, we assign a cumulative ordered spike-and-slab (COSS) prior to the column-specific variances of the item loading matrix. This prior induces increasing shrinkage across latent dimensions, allowing redundant dimensions to be shrunk toward zero while preserving flexibility for active dimensions. Albert--Chib latent response augmentation is used to handle the ordinal probit likelihood, yielding conditionally Gaussian updates for item loadings and latent traits. These updates are combined with Gibbs updates for threshold and shrinkage parameters in an efficient adaptive sampler. Simulation studies evaluate the proposed method in terms of dimension recovery, parameter estimation accuracy, and computational efficiency, with comparisons to conventional fixed-dimensional estimation and model selection procedures. The results show that the proposed approach accurately recovers the latent structure while avoiding repeated model fitting over multiple candidate dimensions. We further illustrate the method using real psychological assessment data, demonstrating its practical utility for uncovering interpretable latent structures in ordinal item responses.

Traditional benchmarks for LLMs primarily rely on static datasets and objective scoring metrics, which often fail to capture differences in response quality when multiple answers are acceptable. In such settings, correctness alone is insufficient to distinguish between responses that vary in clarity, completeness, and usefulness. This paper introduces a consensus-based evaluation framework that measures relative preference among model-generated responses rather than absolute correctness. Instead of evaluating outputs against a fixed ground truth, we assess how a panel of diverse LLMs ranks anonymized candidate responses to the same prompt. This approach treats aggregate inter-model agreement as a proxy for perceived response quality under blind conditions. We conduct a controlled study using five state-of-the-art LLMs across multiple domains, including programming, general knowledge, safety, logical reasoning, and mathematics. Each model generates responses and independently ranks peer outputs through a structured voting process. Scores are aggregated into a Relative Intelligence Index (RII), representing how frequently a model's responses are preferred by other models. Our findings reveal consistent preference patterns across domains, with certain models more frequently ranked highly by their peers. However, we emphasize that these results reflect inter-model preference alignment rather than objective correctness or human judgment. This framework provides a scalable, model-driven method for comparative evaluation, offering an alternative perspective on response quality in scenarios where multiple valid answers exist. While not directly aligned with human evaluation, prior work suggests that aggregated model preferences can partially correlate with human judgments, motivating this as a proxy signal.

AI benchmarks increasingly leverage item-level statistical models, particularly item response theory (IRT), to estimate model capabilities, rank systems, select informative examples, and diagnose benchmark quality. However, AI benchmark data often departs from the data regime of human testing, for which standard IRT estimation tools were originally developed: benchmarks typically involve fewer evaluated models, far more items, and capability distributions that may be skewed, clustered, or multimodal. We examine how these regime mismatches challenge the reliability of IRT modeling for AI evaluation. Using item parameters and capability distributions derived from six widely used LLM benchmarks, we simulate response matrices under three common IRT models and compare four estimation tools used in recent benchmark studies: marginal maximum likelihood, Markov chain Monte Carlo, variational inference, and a neural pseudo-Siamese estimator. Across 18,000 simulation conditions, we systematically evaluate computational feasibility, scalability, and the reliability of IRT inferences about model rankings, predicted performance, and item characteristics. Results show that classical estimators can become infeasible in large benchmark settings, whereas scalable estimators can produce unreliable item-level and ranking inferences with small or nonnormally distributed model sets. This study identifies when latent trait models reliably support or risk distorting AI benchmarking claims, and what sample sizes and diagnostics are needed for trustworthy use.

All-in-one image restoration aims to recover clean images degraded by multiple corruption types using a single unified model. Existing methods typically rely on image-level prompts or shared guidance to handle diverse degradations. However, such a paradigm becomes inadequate when degradations are spatially heterogeneous or even coexist in mixed forms within a single image. Yet spatially adaptive guidance alone is not sufficient, since accurate restoration also requires each spatial query to reliably aggregate complementary information from local neighborhoods and global contexts. To this end, we propose QuReC, a unified framework for all-in-one image restoration. QuReC consists of a Degradation-Guided Query Reconstruction Module (DQRM) and a Local-Global Response Calibration Module (LGRCM). Specifically, DQRM matches each spatial query against a degradation prototype space to reconstruct a query-specific degradation-aware representation, thereby providing fine-grained spatially adaptive restoration guidance. To further stabilize this query-wise matching process, we introduce a weakly supervised prototype matching learning strategy to improve optimization stability and degradation semantic consistency. Meanwhile, LGRCM performs local-global dual-branch aggregation and calibrates the aggregated responses with learnable priors, improving the reliability of feature aggregation and the coordination between local detail modeling and global context modeling. Extensive experiments demonstrate that QuReC achieves superior performance on multiple all-in-one image restoration benchmarks. The code is released at https://github.com/zhoushen1/QuReC.

Aphasia following stroke commonly produces systematic naming errors with characteristic profiles, but whether general-purpose language models not designed for clinical simulation can reproduce these patterns remains untested. We investigated (1) whether lesions or controlled perturbations to a multimodal language model can reproduce different types of errors in picture naming, and (2) whether the framework can reproduce the complete error profile of individual persons with aphasia (PWAs). Using LLaVA 1.6, we evaluated perturbation configurations that varied the layer, proportion, and amount of noise applied to model units. We examined 278 PWAs on the Philadelphia Naming Test, classifying responses into seven categories using a validated neural classifier. Six of seven response categories (correct, semantic, mixed, unrelated, neologism, no response errors) emerged at clinically-comparable proportions across distinct parameter space regions, with formal paraphasia being the exception. Searching the perturbation space revealed configurations that reproduced the individual error profile in at least six of seven categories for 97.8% of PWAs and in all seven categories for 79.5% of PWAs. Monte Carlo baselines confirmed that this matching reflects joint inter-category structure rather than marginal overlap. These results establish a quantitative framework for reproducing individual aphasic error patterns in picture naming. They suggest the potential for language models to serve as digital twins of individuals with post-stroke aphasia.

Aggregate reliability estimates can obscure heterogeneity in measurement-design burden across response conditions, so a single G- or D-study may mischaracterize a design's adequacy for particular strata. This study introduces a conditional generalizability framework with three components. First, automated scoring configurations -- the encoder architectures and scoring-head families admissible within a fixed pipeline -- are treated as a universe of admissible measurement conditions rather than incidental modeling choices. Second, analytical D-study projections are compared with empirical configuration sweeps over a finite scoring pool, yielding two estimands of design adequacy whose agreement or divergence diagnoses the realized configuration universe. Third, evidence is conditioned on entropy-defined response strata, treating entropy as an operational stratification variable, not a construct claim about writing quality. Whereas recent generalizability-theory extensions address AI-generated item variants on the response side, this framework addresses the analogous scoring-side problem: AI-mediated scoring configurations. Demonstrated with automated essay scoring of timed L2 writing, the realized design was dependable in aggregate (Phi approx 0.76). Re-estimated within entropy strata, dependability stayed high but declined modestly and robustly (Phi = 0.88, 0.87, 0.84) -- a gradient implying different decision-study requirements, the highest-entropy stratum requiring the most crossed conditions. The framework offers a portable workflow for evaluating nonuniform dependability.

In this paper, we propose a training strategy for confidence estimation in omnidirectional stereo, targeting the ambiguous matches that frequently occur in wide-baseline setups. Reinterpreting the matching responses produced by the 3D encoder decoder block, we show that their expectation values provide intrinsic confidence signals. Building on this, our method directly penalizes ambiguous responses without auxiliary heads, multi-pass inference, or additional modules, resulting in more efficient and generalized predictions. Beyond confidence, we introduce swept feature volume resampling, where response features produced by 3D CNNs are resampled using regressed positive matching indices and then processed by 2D CNNs to predict meta-information such as surface normals. This joint learning introduces auxiliary geometric regularization and improves depth coherence by leveraging additional contextual cues during response aggregation stage. Experimental results demonstrate that our approach enhances both confidence estimation and surface normal prediction while maintaining deployment practicality for autonomous mobility applications.

Recent work has reported Emergent Misalignment (EM), where language models fine-tuned on narrow, domain-specific misaligned datasets abruptly acquire broadly misaligned behavior, alongside evidence that this behavior can be reversed through limited realignment. We systematically study repeated alignment and misalignment cycles using controlled fine-tuning loops while tracking behavioral performance, and LoRA representations throughout training. Although we reproduce EM, we find that both misalignment and realignment are highly sensitive to superficial dataset characteristics, with apparent rapid realignment largely disappearing after controlling for response-length differences. We further find that previously reported mechanistic signatures, including representational phase transitions in LoRA space, do not consistently correlate with behavioral misalignment across training. Our results suggest that current evidence for EM is less robust than previously claimed and highlight the need for evaluation protocols that carefully control for these surface level dataset artifacts to identify the robustness of the EM phenomenon.

Governing equations provide compact descriptions of physical systems, yet the variables in which they are simple are often hidden in high-dimensional measurements. This challenge is sharper for forced systems, whose responses depend on both intrinsic dynamics and time-dependent inputs. Here we introduce FLARE, a forced latent autoencoder for response equations that learns compact response coordinates, identifies sparse input-dependent latent dynamics and decodes equation rollouts to full responses. By estimating latent dimension from data and separating state estimation from external forcing, FLARE enables forecasts to be initialized from past responses and driven by prescribed future inputs. Across known dynamical systems, application-scale forced responses and visual observations, FLARE recovers compact forced dynamics and predicts long-horizon high-dimensional responses under inputs not used for training. By turning learned coordinates into a dynamical interface, FLARE extends equation discovery to systems whose effective states are hidden within complex observations, providing a route for interpretable modelling and prediction of high-dimensional responses in forced dynamical systems.

The multiplicative repetition penalty shipped across the LLM inference ecosystem (HuggingFace, vLLM, llama$.$cpp, and a dozen further engines) branches on the sign of each raw logit (divide positives by theta, multiply negatives). But the softmax is unchanged by adding a constant to every logit, so a model's logit zero-point is arbitrary (a gauge choice), and the sign-branch reads it. Two measurable consequences follow. (1) The penalty is not well-defined: re-centering a model's logits by a constant is a provable no-op at theta=1, yet at a routine theta=1.3 it changes 58-96% of greedy tokens, while subtractive and normalized penalties change none; real checkpoints sit at widely different zero-points, so a fixed repetition_penalty is a different operation on every model. (2) It corrupts structured output: on 200 real-world JSON schemas, theta=1.3 drops the rate of valid, schema-conformant output from 97% to 23%. Applying the penalty to normalized log-probabilities instead of raw logits removes the gauge dependence by construction; HuggingFace's beam search has applied its processor chain, penalty included, to log-probabilities since at least v4.0.0, so repetition_penalty already names two different operators depending on decoding strategy. Because equal theta is not equal strength across the two operators, we also compare them at matched suppression, calibrated per model by search: there the normalized operator is statistically no worse on any quality metric measured, but on four of seven models it cannot match the raw operator's suppression at theta >= 1.15, and on six of seven at theta=1.3, the setting where the corruption was measured. This note gives the mechanism, the measurements (five models up to 7B; two code models; both effects replicated inside vLLM and llama$.$cpp through their own samplers), the per-model calibration map, and the normalized variant.

This paper reports an empirical study evaluating the relevance of several RAG metrics. The experiment is based on a question-answering dataset created by human annotators from business data. The generated responses and retrieved spans of a RAG system are scored using evaluation metrics from four libraries (Ragas, DeepEval, RAGChecker, Opik). These metrics are compared to scores given by two evaluators, as well as to standard metrics such as recall. An analysis of correlations is conducted. Finally, we highlight certain limitations of our methodology, compare it to those used in the literature, and suggest some avenues for future research. This paper is an English translation of a paper originally published in the French-speaking workshop EvalLLM (Brabant, 2026).

The remarkable performance of large language models (LLMs) in linguistic tasks underscores an urgent need for comprehensive evaluation of their response quality. Prevailing methods, often confined to singular dimensions, fall short of capturing the full spectrum of model capabilities. This study introduces a multifactor scoring paradigm, integrating accuracy, conciseness, factual consistency, readability, and coherence, complemented by a graphical user interface (GUI) for visualizing outcomes. Evaluations on the TruthfulQA dataset unveil mainstream LLMs' strengths in reasoning tasks (peaking at a composite score of 0.6104) alongside pervasive limitations in navigating complex facts and ambiguities. Transcending the narrow lens of traditional metrics, this framework offers a transparent, adaptable avenue to illuminate model potential and deficiencies. Though presently focused on English tasks, its horizons beckon toward multilingual domains. This work carves a novel path for knowledge engineering and model refinement.