AI NEWS JOURNAL

AI, IN PERSPECTIVE

Search AI news

Search headlines, summaries and articles to connect the dots.

2026-11-04

arXiv cs.LG13:00

Does the Trust Boundary Hold over Time? A Long-Horizon Benchmark for Receiver-Directed Federated Learning

arXiv:2611.03724v1 Announce Type: new Abstract: In receiver-directed federated learning, each site draws a trust boundary. For every source and model component it judges, on its own data, whether the transfer helps (admit), harms (reject) or cannot yet be told (abstain). We ask whether this boundary holds over hundreds of rounds. Our benchmark has a federation of thirteen semi-synthetic sites, an exact oracle for every evaluated transfer, paired seeds, an attack protocol, fourteen methods over 600 rounds and Qwen3-1.7B LoRA sites over 500. The answer depends on which boundary is measured. Across 20 paired CPU seeds, a gate judging against a fixed reference better identified transfers useful relative to the site's starting model. Yet it admitted 29% of transfers harmful to the model the site now has. A gate judging against the current model admitted 2% of harmful transfers on that target in the CPU tier, but missed many transfers useful to the starting model. Two short propositions explain these distinct failures. We also evaluate a held out audit that leaves gate verdicts intact. After training, it decides whether each final model is deployed or the site falls back to local training, and reports admissions that have gone stale. With independent audit data its guarantees hold after any training history. At 600 rounds the audit deployed 65 to 72% of models, none with negative benefit on its target, and kept 98% of the gain. No stale report had the wrong sign in our runs. With 20,000 audit observations, a variance-adaptive bound found 99% of stale transfers that cost at least 0.01 in log loss. A common-target Qwen3 comparison on 60 selected transfers shows the same tension over 100 rounds. With 38% adversarial sites, FedFomo and CS-Gate were the only evaluated methods with positive mean gain in every honest group.

arXiv cs.AI13:00

El Agente F\'armaco: A Scalable AI Scientist for Drug Discovery

arXiv:2611.03832v1 Announce Type: new Abstract: Computational drug discovery offers a growing range of tools for proposing and evaluating molecules, but combining them and deciding what to do next still requires substantial expert effort. Here, we introduce El Agente F\'armaco, an AI scientist that uses modern large language model (LLM) technology to turn discovery goals expressed in everyday language into complex, scalable computational screening workflows. The agentic harness selects scientific tools, reviews their results, and adjusts subsequent steps to the target and the user's objectives. Our approach centers on a shared scientific memory: F\'armaco Atlas, which stores molecules, structures, and results together with their provenance -- a record of how they were obtained. The agent uses targeted database queries to retrieve the evidence needed for each decision. A single request can launch calculations for an entire collection of molecules, without requiring a separate LLM decision for every compound. This design limits the information the model must process as screening collections grow, while keeping the underlying results available for inspection and further analysis. The integrated tools support molecule generation, prediction of binding and molecular properties, and searches for compounds available for purchase. In a retrospective comparison with an expert-led campaign, the agent proposed chemically distinct candidates with competitive docking scores and higher calculated drug-likeness under explicitly imposed selection criteria. In a 22-task benchmark against plain tool calls and Codex, the agent achieved the highest observed mean model-judged score, with lower median cumulative token use and peak context. F\'armaco provides a foundation for flexible, scalable drug discovery workflows that connect scientific reasoning with an enduring, traceable record of computational work.

arXiv cs.AI13:00

EMBERS: Empirically Motivated Robustness Testing of Deep Research Systems via Defect-Guided Fuzzing

arXiv:2611.03723v1 Announce Type: new Abstract: Deep Research (DR) systems extend large language models into autonomous agents that decompose complex information needs, iteratively search the web, and synthesize evidence to answer challenging questions. However, their multi-stage execution may introduce robustness risks, as variations in user questions or intermediate search inputs may propagate across research stages and lead to incorrect answers. Existing evaluations focus on task-solving capabilities or robustness against external evidence disturbances, leaving intrinsic weaknesses in DR execution insufficiently explored. Thus, we conduct an empirical study of 1,932 execution trajectories from three representative DR systems and identify nine recurring defect categories across subquestion formulation, online search, and final reasoning. Guided by these findings, we propose EMBERS, an empirically motivated, defect-guided fuzzing framework for robustness testing of DR systems. Starting from correctly answered seed questions, EMBERS generates stage-targeted test inputs using defect-guided mutation, web-search-assisted fidelity assurance, and diversity-aware seed maintenance. Experiments across 16 framework--model--dataset settings show that EMBERS achieves 3.21 times the failure-discovery throughput of the strongest baseline under identical testing budgets. Human and LLM-assisted evaluations demonstrate that most generated inputs preserve the original information need and necessary contextual constraints. Moreover, model fine-tuning experiments show consistent-answer performance improvement, highlighting their value for robustness testing and model improvement.

arXiv cs.DC13:00

EvoCache: Accelerating Browser Large Language Model Delivery with Public Cache Anchors

arXiv:2611.04638v1 Announce Type: new Abstract: Large Language Models (LLMs) can run directly in a browser, but first use typically requires downloading model files that are hundreds of MiB or even several GiB in size. The same LLM is often repackaged by different publishers; these files share substantial content, yet byte differences prevent whole-file caches from reusing them. We propose EvoCache, a browser-managed LLM delivery system. EvoCache uses public cache anchors that are stored in advance by the browser and independent of browsing history to reconstruct losslessly the publisher version requested by a site. A site specifies only the target, while the browser selects the anchor and performs reconstruction, keeping both hidden from the site. An audit of 1,000 public repositories finds that 61 of the 100 most popular LLMs are offered by multiple publishers and that 98.1% of overlapping versions cannot be reused by whole-file caching. Across 23 cross-publisher model-package pairs, EvoCache achieves a median traffic reduction of 77.3% and outperforms every comparator. Across 240 real end-to-end browser experiments, EvoCache reduces network traffic by 78.9% and page-ready time by 64.2% relative to ordinary full downloads. For cached data left by other sites, EvoCache also protects users' cross-site cache privacy by using the same return time for cache hits and misses.

arXiv cs.DC13:00

FenixWalk: Adaptive Sampling for Large-Scale Dynamic Graphs on Heterogeneous Architectures

arXiv:2611.04683v1 Announce Type: new Abstract: Random walks are fundamental to graph analytics, yet executing them efficiently on large, evolving graphs remains challenging. Existing approaches must balance sampling efficiency, update overhead, and memory consumption. Sampling structures optimized for throughput often incur substantial memory overhead and costly maintenance, whereas lightweight alternatives can suffer from high sampling latency on high-degree vertices or skewed weight distributions. These challenges become more pronounced on heterogeneous CPU--GPU platforms, where host memory provides capacity but GPU memory limits the amount of graph data that can benefit from parallel acceleration. We present FenixWalk, a memory-efficient random-walk system that combines unified in-place sampling with selective GPU acceleration for large-scale dynamic graphs. At its core, FenixWalk employs an in-place Fenwick representation that supports both rejection sampling and inverse-transform sampling without duplicating edge-level weight storage. This representation enables expected O(1) weight lookup for rejection sampling, together with O(log d) prefix sampling and updates. To exploit heterogeneous hardware efficiently, FenixWalk maintains the complete graph and its sampling structures in host memory while selectively placing frequently reused vertex-local samplers on the GPU. By transferring compact walker states instead of repeatedly migrating graph partitions, the system reduces cross-device communication and makes effective use of limited GPU memory. Experiments on large real-world graphs and representative random-walk workloads demonstrate that FenixWalk achieves average speedups ranging from 2.06x to 400.57x over existing random-walk systems.

arXiv cs.CV13:00

From Visual Matching to Textual Conditioning: Cross-Domain Few-Shot Segmentation via Reference Inversion

arXiv:2611.04084v1 Announce Type: new Abstract: Cross-Domain Few-Shot Segmentation (CD-FSS) aims to segment objects from unseen domains using only a few annotated references. Recent Segment Anything Model (SAM)-based methods have advanced CD-FSS, yet their reference-target visual matching is highly sensitive to target appearance variations. SAM3 introduces text-conditioned segmentation, enabling natural-language descriptions to guide segmentation across target images, but directly using the category name as the condition still yields unstable performance under severe domain shifts. We propose Reference Inversion for SAM3 (RI-SAM3) that leverages SAM3's large-scale vision-language pre-training to learn suitable textual tokens for each target category. RI-SAM3 comprises three core designs: (i) Bi-level Reference Inversion that learns complementary tokens before and after the frozen text encoder: the former remain anchored in SAM3's text space while the latter represent the category more flexibly; (ii) Layer-adaptive Textual Conditioning combines these tokens differently as object queries evolve across decoder layers; and (iii) Presence-aware Confidence Supervision that jointly supervises query confidence and target presence for reliable predictions. With SAM3 frozen, RI-SAM3 optimizes around 12K parameters only per category and it efficiently segments each target in a single forward pass with reusable textual conditioning without per-sample reference-target visual matching. Extensive experiments over multiple widely adopted benchmarks show that RI-SAM3 achieves superior performance consistently in CD-FSS tasks, together with strong robustness, efficiency, and generalization.

arXiv cs.DC13:00

GC-MoE: Overlapping Expert-Parallel Token Dispatch with FP8/FP4 Grouped GEMM via CUDA Green Contexts

arXiv:2611.04222v1 Announce Type: new Abstract: Under expert parallelism, mixture-of-experts (MoE) layers pay all-to-all token dispatch, grouped GEMMs, and a combine step in every layer. In decode each rank holds only 1-256 tokens: dispatch latency and weight traffic dominate while tensor-core utilization is inherently low. In fused operators such as DeepGEMM's Mega MoE, dispatch and computation overlap only by sharing every SM: dispatch warps compete with TMA/MMA and epilogue warps for registers, shared memory, and issue slots; every phase shares one CTA resource budget, synchronization spans the whole SM array, and every replay ends with a workspace-clean pass and a cross-rank barrier. We present GC-MoE, an open-source FP8/FP4 MoE inference operator for NVIDIA Hopper and Blackwell GPUs that uses CUDA Green Contexts to split the device into a small communication partition and a large compute partition. Each role gets its own kernel, resource budget, and synchronization domain, so dispatch and GEMM tune independently. A three-node CUDA Graph (four on Hopper with fused shared experts) launches the dispatch and grouped-GEMM kernels as concurrent roots; a token-ring handshake starts computation on the first filled ring blocks, and double-buffered receive counters remove the trailing cross-rank barrier between replays. The compute partition runs operand-swapped tcgen05 block-scaled MMA on Blackwell and register-A WGMMA with in-register FP4 decoding on Hopper, with fused SwiGLU, requantization, and optional shared experts. On 8 H20 GPUs GC-MoE accelerates DeepSeek-V4.1-Flash decode by 1.188x on average on both the routed and shared-expert suites (peaks 1.296x and 1.353x); on 8 B300 GPUs it accelerates GLM-5.2 and DeepSeek-V4.1-Flash by 1.106x (1.123x with fused shared experts, peak 1.209x), beating Mega MoE at all 54 measured points at matched accuracy. GC-MoE reserves only 20 of 148 SMs on B300 for dispatch.

arXiv cs.CV13:00

Geometric Refinement and Distributional Alignment for Sim2Real Roadside LiDAR Detection

arXiv:2611.03936v1 Announce Type: new Abstract: We describe our approach to the LUMPI track of the UrbanTwin Sim2Real LiDAR Challenge, in which a 3D object detector must be trained exclusively on synthetic data and evaluated on unseen real roadside LiDAR, while the synthetic point clouds are simultaneously scored for distributional realism. Our final entry scores 0.4793 combined, from a detection mAP of 0.1333 and a realism score of 0.9126. Our contribution is not a new architecture. Working from a PointPillars baseline, we obtained most of our gains from three sources. First, a diagnosis that small object failure in roadside LiDAR is a localisation problem rather than a detection problem, which led to a parameter free geometric refinement worth 86 percent relative on pedestrian AP. Second, an audit of our own post processing which revealed that a shipped rule was destroying valid predictions, the repair of which produced our single largest detection gain at 335 percent relative on the Van class. Third, treating the realism metrics analytically rather than as a black box, which showed that one of the four components is a closed form function of nine summary statistics. We also report an experimental methodology that we believe is the more transferable result. Because mean AP decomposes into independent per class terms, a single submission can carry several mutually controlled experiments, and a submission assembled entirely from previously measured components has a score that is known before it is sent. We predicted our final score to four decimal places prior to submission. We report negative results at equal length, and give a full account of data provenance and compliance in Section 8, including a full account of every mechanism contributing to the submitted score.

arXiv cs.LG13:00

Latent Cluster Separation as a Regularizer for Lifelong Anomaly Detection

arXiv:2611.03652v1 Announce Type: new Abstract: Lifelong (continual) anomaly detection systems must learn a sequence of tasks without forgetting how to recognize anomaly types seen earlier -- a property known as backward transfer. A common way to try to improve backward transfer is to reshape a model's internal representation around learnable concept centroids, typically coupled with a new way of scoring anomalies based on distance to those centroids. Because these two design choices are usually introduced together, it is unclear which one is actually responsible for any performance gain. In this paper, we unbundle the two mechanisms within a VAE-based lifelong anomaly detector and study them separately. We find that used purely as a training-time regularizer -- while leaving anomaly scoring unchanged -- latent cluster separation can improve detection quality and backward transfer, though whether it outperforms the alternative scoring approach depends on the dataset's underlying representation. Switching to distance-based scoring is instead where most of the instability commonly associated with this class of method actually originates; we identify and explain the underlying mechanism. This benefit depends on the dataset's underlying representation, and we report this as practical guidance for choosing between the two design choices, rather than as a list of caveats.

arXiv cs.DC13:00

LiquidCP: Decoupling Attention Tasks from Communication and Kernel Boundaries in Context Parallelism

arXiv:2611.04401v1 Announce Type: new Abstract: Context parallelism (CP) partitions long sequences across GPUs for LLM training and inference. For variable-length training and full prefill, attention-balanced placement can leave token loads uneven, while ring execution repeats overhead at each computation and communication step. In stage-serialized decode CP (DCP), long-running tasks delay dependent work and leave substantial GPU hardware idle. We propose LiquidCP, which unifies training and inference through a common model of Transfer, Compute, and Reduction tasks with tile-level dependencies, decoupling attention tasks from communication and kernel boundaries. For training and full prefill, MegaRing incorporates BuddyRing placement to balance attention and token loads, and coalesces ready KV regions during forward attention to amortize overhead. For decoding and mixed batches, MegaDCP selects history-KV splits by jointly modeling the completion of attention and its dependent reductions. A shared persistent GPU-resident engine executes both designs. Compared with the fastest baseline per configuration on eight H100 GPUs, MegaRing achieves geometric-mean core-attention speedups of $1.23\times$ forward and $1.18\times$ backward, and accelerates a Transformer layer by $1.08$--$1.16\times$. MegaDCP achieves geometric-mean core-attention speedups of $1.20\times$ for decode-only and $1.35\times$ for mixed batches. In vLLM serving with KV prefix cache, it achieves an average time-to-first-token speedup of $1.22\times$.

arXiv cs.AI13:00

LOLPlay: Benchmarking Real-Time Visual Agents in League of Legends

arXiv:2611.03685v1 Announce Type: new Abstract: Visual agents need to coordinate reasoning and action while their environment evolves during inference. We introduce LOLPlay, a benchmark in native \emph{League of Legends} with 36 champion-specific tasks spanning resource management, combat, cooperation, and objectives. Agents interact through screenshots and mouse--keyboard actions; automated initialization and 82 executable subgoal checks distinguish partial progress from completion. Across 324 task evaluations of three models at three reasoning-effort settings, GPT-6 Astra achieves the highest task success rate of 25.0\% at low and medium effort, and the highest subgoal success rate of 40.3\% at medium effort. However, high effort improves neither metric over low effort for any model. For Astra, longer requests leave fewer interaction rounds, suggesting a cost to additional reasoning in real-time control. Trajectories reveal incomplete subgoal composition and actions executing after the observed scene changes, motivating agents that combine visual reasoning with timely action.

arXiv cs.AI13:00

MCPGuard: Finding Conformance Gaps in MCP SDKs via Dual-Track Analysis and Dynamic Adjudication

arXiv:2611.03662v1 Announce Type: new Abstract: The Model Context Protocol (MCP) connects large language model agents to external tools and services through official SDKs that implement message validation and session logic. Specification constraints omitted by these SDKs may propagate to downstream applications, highlighting the importance of SDK-level conformance testing. Yet existing MCP conformance tools mainly exercise running endpoints and reveal little about whether SDK models enforce specification requirements. We present MCPGuard, an automated LLM-based framework for detecting specification-to-SDK conformance gaps. Its structural track aligns JSON Schema constraints with normalized SDK models and produces constraint-specific evidence. Its behavioral track extracts normative prose rules, retrieves bounded function-level slices, and audits them with LLMs for candidate omissions. An evidence-aware dynamic adjudicator then inspects complete source context and builds PoCs to validate candidates. A shared kernel with language-specific adapters enables analysis across SDKs. On the official Python 1.26.0 and TypeScript 1.29.0 SDKs, MCPGuard confirms 13 and 19 findings, respectively, none of which are exposed by two black-box baselines. Tests on open MCP servers further show that these gaps can systematically propagate downstream. By making SDK conformance of specification constraints directly testable, MCPGuard complements conventional endpoint-oriented conformance suites.

arXiv cs.LG13:00

Minimal Superior Penalties: Characterising and Calibrating Decision-Aware Proper Scoring Rules

arXiv:2611.03734v1 Announce Type: new Abstract: Training a classifier means letting a validation metric decide which model is kept and when training stops. A natural requirement on that metric is superiority: every correctly classified forecast should score better than every incorrectly classified one, however close the underlying probabilities are. Neither the Brier score nor the logarithmic score satisfies this on its own. We show that superiority is governed by the separation between two endpoints -- the supremum of the score on correct predictions and the infimum on incorrect predictions -- and derive both in closed form for each score. Their difference is the infimal penalty, the greatest lower bound on the penalty needed to restore superiority. Prior work restores superiority with a penalty calibrated from the first endpoint alone; for the Brier base, our infimal penalty is smaller than that published penalty by a factor ranging from 4.00 at three classes to 2.11 at twenty. We further embed the published and infimal penalties in a one-parameter family and show that, under a single-valued tie convention, every member of the family is strictly proper. In a pre-specified two-dataset case study, mean selected-model macro-F1 rises with the penalty in the reported ten-class regimes, though not distinguishably from noise once seed/selection-mode nesting is accounted for; a fifth regime at forty-seven classes does not reproduce even this descriptive pattern. The intended no-superiority baseline is algebraically a rescaling of the family under test, so the experiment cannot attribute the observed pattern specifically to superiority. When superiority itself is the objective, any penalty strictly above the infimal value suffices; choosing a larger penalty is a separate calibration trade-off rather than a requirement for the guarantee.

arXiv cs.DC13:00

Minimizing Bottleneck Overload Risk in Online Edge Task Placement under Workload Uncertainty

arXiv:2611.04425v1 Announce Type: new Abstract: In edge computing, tasks must be placed on servers of finite capacity before workloads are realized, often with only workload means and variances known. An assignment balanced on expected load can still leave a server with more fluctuations than its residual capacity can absorb. Per-server risk constraints can enforce a prescribed overload tolerance but do not distinguish the assignments that meet it by how close the riskiest server comes to the limit. We show that, for independent tasks, placement preserves a period's total variance and total residual capacity, and decides only how fluctuation, measured by standard deviation, is paired with residual capacity across servers. The risk ratio, the ratio of fluctuation to residual capacity, parameterizes a server's worst-case overload probability over all aggregate workload distributions with given mean and variance. We minimize the largest risk ratio, attained at the risk bottleneck. Risk-Bottleneck Placement (RBP) places tasks in descending order of standard deviation, each on the server with the smallest post-assignment risk ratio, and refines the assignment with relocations and swaps within a finite per-period online budget. On small instances solved exactly by enumerating every assignment, minimizing the maximum expected utilization leaves the largest risk ratio 25.66-29.19% above its exact minimum on average, even with risk-minimizing tie-breaking, and the corresponding gap for RBP is 1.77-2.64%. In trace-derived multi-period simulations at two workload pressures, RBP reduces the worst-server overload probability by 26.0% and 18.0% relative to the best baseline and remains the lowest at every tested pressure and scale.

arXiv cs.AI13:00

Predictive gains, causal losses: why interventional competence cannot be learned from data alone

arXiv:2611.03861v1 Announce Type: new Abstract: Retaining predictive gains can require persistent errors in a model's response to interventions. For deterministic predictors with finite second moments under squared-error loss, any observational advantage over the best predictor using only an outcome's direct causes requires dependence on other inputs, regardless of the learning rule. When admissible interventions vary inputs while preserving the worldly outcome, this dependence yields a positive lower bound on root-mean-square intervention-response error. We identify conditions under which population gradient updates introduce this dependence from a parent-only baseline or damage a previously correct mechanism through shared parameters while improving other predictions. These dependencies can also make model behavior vary with how the same intervention is realized. We characterize interventional alignment through a common output mapping across internal states and physical processes realizing the same intervention target and relevant context. Correction is limited by the causal hypotheses currently admitted within inquiry. The relevant hypothesis class comprises these causal alternatives, rather than the function class the computational system could in principle represent. For predictions restricted to these alternatives, interventional mean error cannot fall below the best approximation they permit, whether evidence is observational or interventional. Reducing error below a positive limit requires admitting better hypotheses, which may entail revising the causal variables, mechanism structure, or intervention interpretation. Interventional criticism constrains replacements without determining them. We conjecture that reliably acquiring and maintaining interventional competence when causal commitments fail requires autonomous participation in proposing alternatives, testing their consequences, and revising them in response to error.

arXiv cs.LG13:00

Probabilistic Steering of Generative Priors for Lab-in-the-Loop Antibody Optimization

arXiv:2611.03844v1 Announce Type: new Abstract: Therapeutic antibody development is often framed as comprising two discrete stages: initial discovery of novel binders followed by iterative optimization to refine these hits into therapeutic-grade molecules. Models and methods often reflect these stages, with generative structure models applied to discovery and probabilistic black-box optimization methods applied to refinement. Here we propose to unify these stages by smoothly transitioning a generative discovery model into iterative optimization, allowing the handoff between the structure-based prior and supervision from wet lab data to be guided by the posterior uncertainty of a probabilistic surrogate. We show that this yields improvements in optimization efficiency in the low-N regime, as well as improved diversity amongst the optimized molecules.

arXiv cs.AI13:00

Reasoning is topologically simple

arXiv:2611.03739v1 Announce Type: new Abstract: We formulate and investigate the hypothesis that reasoning text has simpler co-occurrence topology than non-reasoning text, and that higher-quality reasoning has simpler co-occurrence topology than lower-quality reasoning. We discuss and test applications to AI safety and AI evaluation.

arXiv cs.CL13:00

Rethinking Next-Visit Diagnosis Prediction Beyond Aggregate Metrics: Retrieval--Memory Multi-Agent Adjudication

arXiv:2611.03826v1 Announce Type: new Abstract: Next-visit diagnosis prediction forecasts the diagnoses documented at a patient's next encounter, turning retrospective electronic health records into prospective clinical prediction. However, conventional aggregate evaluation obscures its multifaceted nature: distributionally, models must handle variable-size diagnosis sets and long-tailed labels; temporally, they must distinguish persistent, newly recorded, and recurrent diagnoses; and both vary with observed history depth. We introduce an evaluation framework along these dimensions and apply it to rule-based heuristics, LLM-based prompting, retrieval, memory, and multi-agent approaches. This reveals that repeating the last visit outperforms most LLM-based approaches overall despite predicting no new or recurrent diagnoses, and that retrieval and memory exhibit distinct profiles whose relative strengths vary across temporal transitions and backbones. Motivated by this, we propose LONGINUS, a retrieval-memory multi-agent adjudication framework combining evidence retrieved from patient records with structured memory of the patient trajectory. Using cross-view agreement as a reliability heuristic, it preserves shared predictions and adjudicates disagreements without revealing their source. Across MIMIC-III and MIMIC-IV with four LLM backbones, LONGINUS achieves the best transition-specific performance among LLM-based methods in nearly all settings while remaining competitive with them on aggregate metrics.

arXiv cs.CL13:00

Same Rule, Different Verdicts: Rubric Language and Clause Structure in LLM Judges for Korean Support Dialogues

arXiv:2611.04091v1 Announce Type: new Abstract: Services that operate in Korean often write their evaluation rubrics in Korean and hand them to an LLM judge. We test whether a judge's verdict depends on the language of the rubric and on how a rubric item is phrased. We write 12 criteria for support dialogues in four clause structures (affirmative, negated, conditional, exception) such that two pairs of structures have the same meaning, and we give each item in Korean and in English. Six low-cost models from three providers, plus one reference model, judge 60 translated ESConv dialogues and 144 constructed boundary cases under every combination (51,744 requests across the seven models). The analysis plan was fixed and committed before data collection. Cross-provider agreement is lower with Korean rubrics than with English rubrics (Cohen's kappa 0.687 vs. 0.745; difference -0.058, 95% CI [-0.072,-0.045]). Disagreement concentrates in negated, conditional, and exception items (odds ratio 5.6 relative to affirmative items). A single model changes its verdict between two items with the same meaning in 11-15% of cases. Two hypotheses are not supported: models from the same provider do not agree more than models from different providers, and adding paired examples does not reduce the language gap. We measure consistency only; no human labels are used, and all materials were drafted by an LLM.

arXiv cs.CL13:00

Say Less, Mean Better: Training-Free Verbosity Control for Language Models

arXiv:2611.04107v1 Announce Type: new Abstract: Large language models (LLMs) often generate responses longer than necessary, increasing inference latency and computational cost. While prompting can encourage conciseness, controlling output length without compromising response completeness remains challenging. Aggressive early stopping can reduce generation costs but risks terminating responses prematurely. We introduce MinimalLM, a training-free framework for reliable verbosity control that combines uncertainty-aware, length-dependent stop-token adjustment with structural completeness guardrails. MinimalLM separates the decision of when to stop from whether stopping is structurally appropriate, coordinating response-level instructions, decoding-time intervention, and completion checks without modifying model parameters or requiring auxiliary models. We evaluate MinimalLM across three open-source LLMs and 330 questions, producing 5,940 responses across baseline and ablation configurations. MinimalLM reduces generated tokens by 47.6%, accelerates individual responses by 19-68%, and eliminates structurally truncated outputs in the primary evaluation. Despite substantial token savings, it achieves an average quality score of 8.04/10, compared with 8.42/10 for unmodified models and 6.46/10 for length-matched truncation. Evaluations across diverse subject domains (such as healthcare, programming, law, and science) demonstrate robust efficiency gains, while ablations establish the complementary roles of decoding control and completeness constraints. Our findings highlight that efficient generation requires more than producing fewer tokens: it requires controlling when generation ends while preserving response integrity. MinimalLM offers a lightweight, training-free approach toward this goal.