Searcharxiv⌕ Search

arXiv subjects

Yubo Li

Publications and source records attributed to Yubo Li.

At least 19 recordsLinked to original sources
arXiv2026

JEV-as-a-Judge: Accept When Confident, Escalate When Unsure

LLM-as-a-judge scales evaluation, but reasoning judges are slow and costly. We study JEV-as-a-Judge: evaluation with JEV, a decision-only judge that returns label probabilities instead of text, and whose confidence decides whether to accept its verdict or escalate to a reasoning judge. Against sixteen generative and reward-model judges, with blinded human adjudication, JEV comes within three points of GPT-6 wherever a verdict can be read off the text, at 0.36% of its fee and a 0.15-second median latency, and falls behind where the verdict must be derived, as in math, code, and logic. Its confidence marks this boundary. With a threshold frozen in advance, accepting confident verdicts and escalating the rest is 0.9 points more accurate than GPT-6 on 1,610 held-out pairs at 41% of its fee, and in a pre-specified live test on two new workloads the cascade matches GPT-6's accuracy exactly. Confidence routing weakens on style-adversarial pairs and reference-free prose; we close with a simple recipe for validating thresholds locally.

cs.AI↗
arXiv2026

Learning When to Think: Utility-Guided Adaptive Reasoning for Efficient Text Retrieval

Reasoning-conditioned embeddings can improve retrieval by progressively refining semantic constraints. However, additional reasoning may introduce over-generalized or semantically noisy abstractions that weaken retrieval specificity. Existing methods largely supervise reasoning through final retrieval outcomes, providing limited guidance over the refinement process itself. We propose AURA, a framework that directly supervises intermediate reasoning states according to how well they separate relevant documents from hard distractors. AURA further trains a lightweight utility estimator over prefix-conditioned embeddings to adapt reasoning depth at inference time. On the reasoning-intensive BRIGHT benchmark, AURA-4B achieves 34.7 average nDCG@10 while using 2.8 reasoning steps and 96 generated tokens per query. Under matched deployment conditions, adaptive stopping reduces latency from 4.6 to 0.76 seconds relative to full-trajectory reasoning. Results on MS MARCO further suggest that this specialization need not substantially degrade conventional in-domain retrieval.

cs.IR↗
arXiv2026

Reading Right, Acting Wrong: Diagnosing Decision Failures of System One Models in Agent Loops

System One models choose an agent's next action by scoring a supplied set of options in a single pass. This makes them fast and cheap, but when they choose wrongly there is nothing to inspect: did the model misread what the agent has seen, or read it correctly and still act wrongly? We answer behaviorally. On a frozen benchmark of agent decision states from simulated environments, customer-service policies, tool use, web navigation and coding, we pair every next-action question with factual probes about the same state, each answered independently. Across seventeen fast systems and a reasoning model, the answer is consistent: most wrong actions follow correct reading, and the more accurate a system, the more of its errors take this form. These errors are not noise; they gather on the same states across systems of every kind. Where the probed facts provably determine the action, models that read every fact still fail to combine them, and their own answers, composed by code, repair the failure. On real states, what is missing is often not a fact but the next sub-goal.

cs.AI↗
arXiv2026

The Model Says Walk: Measuring whether LLMs Condition on Hidden Constraints

Asked whether to walk or drive to a car wash 50 m away, most language models say walk, forgetting that the car has to be there. Such failures are usually measured by accuracy on questions where the hidden constraint applies. We show that this is misleading: a model can answer these questions correctly without reasoning about the constraint at all, simply by favoring the cautious option. We instead ask whether a model's decision changes when the constraint is removed. We test this at three levels of control: log-probability sweeps on open-weight models, a 500-prompt stress test, and CORE, a new human-validated benchmark of minimal constraint-present/constraint-absent pairs. The picture is consistent. The constraint nudges decisions rather than governing them. Models still miss presence constraints like the car wash, yet elsewhere they apply constraints that are not there. As a result, standard accuracy flatters all ten models we evaluate and reorders their ranking, and prompting fixes that look effective largely vanish under paired scoring. Claims about hidden-constraint reasoning need paired evidence.

cs.CL↗
arXiv2026

Auditing Institutional Heterogeneity for Generative AI in Patient Education: A Large-Scale Study of 102 US Transplant Handbooks

Health systems are rapidly deploying generative-AI assistants that answer patient questions from institution-authored education materials, on the premise that grounding in local content yields consistent guidance. Do the underlying documents themselves agree? We use a structured-output large-language-model judge to audit 1{,}772{,}261 pairwise comparisons across 102 patient-education handbooks from 23 US solid-organ transplant centers, paired with 1{,}115 patient-derived questions (TransplantQA). Four findings bear directly on deployment: (1) same-center cross-organ agreement exceeds cross-center same-organ agreement by $0.024$ in the primary analysis (Holm-adjusted $p=0.011$), with sensitivity to document selection; (2) information gaps concern topics relevant to underrepresented subgroups, with reproductive health a \emph{double jeopardy}: 82\% absence and 86\% judge-rated high significance among divergent/contradictory pairs; (3) judge-derived themes form 991 clusters, with immunosuppression and pregnancy timing among the highest judge-rated priorities; (4) question and observed-coverage features predict high-divergence questions retrospectively (AUC $0.77$). We discuss implications for deploying patient-facing generative AI in transplant care.

cs.CY↗
arXiv2026

Dynamic Agent Skills: A Lifecycle Survey and Taxonomy of Evolving Skill Libraries

Large language model agents increasingly store reusable procedures outside the model. These reusable procedures are often called \emph{skills}: they may be code functions, natural-language instructions, SKILL.md packages, workflow graphs, or learned adapters that a future agent can retrieve and invoke. This taxonomy-driven survey asks how such skill libraries change over time. Across a $124$-paper $2023$--$2026$ audit set, we synthesize dynamic skill systems as \emph{lifecycle-managed, verified, evolving artifact stores}: agents collect evidence from interaction, propose skill updates, verify and admit candidates, organize them for retrieval and composition, repair or prune stale entries, and govern sharing through provenance and rollback. We organize the literature around three survey tools. First, a $\text{six}$-sense taxonomy distinguishes the structurally different artifacts called ``skills'' in current papers. Second, an $\text{eight}$-stage lifecycle architecture identifies the recurring design decisions behind evidence acquisition, proposal, verification/admission, storage, retrieval/composition, maintenance, distillation/portability, and governance. Third, a lightweight skill-record schema and $\text{ten}$-operator vocabulary provide common terms for comparing library updates without elevating them into a separate method contribution. Using this structure, we synthesize evidence-graded patterns with explicit caveats: admission and repair are repeatedly important, verifier quality materially affects skill-aware RL, flat retrieval can degrade as libraries grow, and current benchmarks still under-report library trajectories, usage--utility gaps, and safety surfaces. We close with concrete reporting standards and open problems for evaluating dynamic skills as changing libraries rather than static prompt or tool collections.

cs.AI↗
arXiv2026

Let Language Constrain Geometry: Vision-Language Models as Semantic and Spatial Critics for 3D Generation

Text-to-3D generation has advanced rapidly, yet state-of-the-art models, encompassing both optimization-based and feed-forward architectures, still face two fundamental limitations. First, they struggle with coarse semantic alignment, often failing to capture fine-grained prompt details. Second, they lack robust 3D spatial understanding, leading to geometric inconsistencies and catastrophic failures in part assembly and spatial relationships. To address these challenges, we propose VLM3D, a general framework that repurposes large vision-language models (VLMs) as powerful, differentiable semantic and spatial critics. Our core contribution is a dual-query critic signal derived from the VLM's Yes or No log-odds, which assesses both semantic fidelity and geometric coherence. We demonstrate the generality of this guidance signal across two distinct paradigms: (1) As a reward objective for optimization-based pipelines, VLM3D significantly outperforms existing methods on standard benchmarks. (2) As a test-time guidance module for feed-forward pipelines, it actively steers the iterative sampling process of SOTA native 3D models to correct severe spatial errors. VLM3D establishes a principled and generalizable path to inject the VLM's rich, language-grounded understanding of both semantics and space into diverse 3D generative pipelines.

cs.CV↗
arXiv2026

SkillsBench: Benchmarking How Well Agent Skills Work Across Diverse Tasks

Agent Skills are structured packages of procedural knowledge that augment large language model (LLM) agents at inference time. Despite rapid adoption, there is no standard way to measure whether they actually help. We present SkillsBench, a benchmark whose current inventory contains 87 tasks across 8 domains paired with curated Skills and deterministic verifiers. Our latest aggregate evaluation runs the 87-task benchmark under matched no-Skills and curated-Skills conditions for 18 model-harness configurations. Curated Skills raise the average pass rate from 33.9% to 50.5% (+16.6 percentage points; 25.5% normalized gain), with configuration-level gains ranging from +4.1 to +25.7 pp. Focused Skills with at most three modules outperform larger or exhaustive bundles, and smaller models with Skills can match larger models without them. SkillsBench establishes paired evaluation as the foundation for rigorous measurement of Skill efficacy on agentic, expertise-heavy work.

cs.AI↗
arXiv2026

LLMs Know the Constraint But Do Not Use It: Activation Bottlenecks in Pragmatic Constraint Reasoning

When a salient surface cue competes with an implicit feasibility constraint, LLMs often fail -- but aggregate accuracy conflates genuine constraint inference with conservative defaulting. We formalize the distinction as conditional constraint activation: the constraint is internally encoded (Knowledge) symmetrically across constraint-present and -absent prompts (Symmetry), yet only sometimes routed into the decision (Routing) and repairable by a donor activation (Repair). A quartet diagnostic over 14 models reveals two failure modes; probes on two open weights decode the constraint above $88\%$, yet activation patching repairs one ($+6.4$ nats) and not the other ($-0.07$). On a mitigation frontier, no prompted intervention reaches the repair corner: all inflate conservative bias through a single mediation pathway -- prerequisite mention. Hidden-constraint failure is a routing problem, not a knowledge problem.

cs.CL↗
arXiv2026

Same Question, Different Source, Different Answer: Auditing Source-Dependence in Medical Multi-Source RAG

A retrieval-augmented generation (RAG) system deployed over a multi-author institutional corpus can give a different answer to the same question depending on which source it retrieves -- a failure mode the dominant single-gold-answer paradigm cannot diagnose. We argue that source-dependence is a missing axis of NLP evaluation, and that auditing it means shifting the unit of evaluation from answer correctness to the inter-source relationship. We make this concrete in transplant patient education, where institutional sources demonstrably disagree, releasing three artefacts: TransplantQA, a benchmark of real patient questions, each answered by grounding generation in multiple institutional handbooks as candidate sources; HERO-QA, a hierarchical retrieval strategy that grounds and audits each answer; and a structured-output judge that scores inter-source relationships on a validated 5-label taxonomy. At scale, better retrieval reveals far more disagreement than prior estimates suggested -- understating its prevalence, not its intensity. The framework is domain-agnostic and transfers to legal and educational RAG: measuring source-dependence is a responsibility for deployed multi-source NLP generally.

cs.CL↗
arXiv2026

The Chain Holds, the Answer Folds: Trace-Answer Dissociation in Reasoning Models Under Adversarial Pressure

Reasoning models are evaluated on single-turn benchmarks but deployed in multi-turn dialogue, where users push back on correct answers. Under sustained adversarial pressure we find a previously undocumented failure mode: the chain-of-thought stays factually correct from first turn to last while the emitted answer flips wrong. We call this unfaithful capitulation (UC) and isolate it with a $2\times 2$ latent-versus-behavioral framework that flip-rate metrics and single-turn faithfulness probes both miss. Across three datasets (MT-Consistency, MMLU-Pro, GSM8K), the latent-correct rate at the behavioral flip clusters near 50% in think mode and collapses to 11-15% under no_think -- paired, within-model causal evidence that reasoning creates the gap. Across models the effect tracks the reasoning channel (high in Qwen3-32B and GPT-OSS-20B, low in inline-CoT Gemma-4-31B-it). An independent GPT-4o judge corroborates $86\%$ of UC labels; a token-level probe shows the answer-slot argmax is correct in $84\%$ of UC cells; and a naive trace-anchored defense backfires. We release all trajectories, traces, and judge labels.

cs.AI↗
arXiv2026

PANDO: Efficient Multimodal AI Agents via Online Skill Distillation

Recent advances in multimodal web agents often rely on increased inference-time computation, including rollout search, verifier passes, offline skill discovery, and specialist model stacks. This raises a central question: can a web agent become more efficient as it accumulates experience, rather than more expensive? We first analyze trajectories from VisualWebArena and identify three recurring sources of inefficiency: repeat-action loops, hidden discovery costs, and low prompt-cache reuse. We then introduce PANDO, a single-rollout online skill-distillation framework that maintains a structured Skill Library and combines progress reflection, confidence-based skill demotion, hierarchical routing, visual compression, and cache-aware prompting. On the full set of 910 VisualWebArena tasks, PANDO achieves a 58.3% success rate, outperforming SGV (54.0%) and our WALT reproduction (45.2%), while using 58% fewer tokens than SGV and 61% fewer tokens than WALT, without any pre-evaluation discovery budget. A 300-task ablation further shows that rules and routines provide most of the success gains, while routing, compression, and cache-aware prompting convert the larger skill library into lower marginal token cost. Finally, we introduce three trajectory-level efficiency metrics -- Action Repetition Rate, Step Overhead Ratio, and Prompt Cache Utilization -- to make efficiency visible beyond terminal success.

cs.AI↗
arXiv2026

CONF-KV: Confidence-Aware KV Cache Eviction with Mixed-Precision Storage for Long-Horizon LLM

Long-horizon LLM inference turns the key--value (KV) cache into the dominant GPU memory consumer and makes per-token attention increasingly expensive. Many common eviction policies use static recency windows or historical attention, leaving unused a signal computed on every decoding step: the model's current uncertainty. We introduce CONF-KV, a KV-cache manager that converts the next-token distribution into a scalar confidence score and uses it to choose the per-step cache budget, retaining more context when the model is uncertain and pruning aggressively when it is confident. Within each budget, tokens are ranked by a composite of accumulated attention mass and recency, while a protected recent window preserves local coherence. We combine the policy with blockwise online-softmax attention, mixed FP16/INT8 storage, and a pyramidal per-layer budget variant. Across four model families and generated lengths up to 4K, CONF-KV stays near the footprint of a fixed 512-token sliding window while remaining within 1.5--2.1 perplexity points of full KV. On Needle-in-a-Haystack up to 32K tokens, CONF-KV reaches 91.4% retrieval accuracy versus 53.8% for sliding windows and 80.6% for H2O; on 75 VisualWebArena tasks it retains 95.3% of full-KV success at 2.8 times lower peak memory.

cs.LG↗
arXiv2026

Beyond Single-Turn: A Survey on Multi-Turn Interactions with Large Language Models

Recent advances in large language models (LLMs) have substantially improved single-turn task performance, yet real-world applications increasingly demand sophisticated multi-turn interactions. This survey provides a comprehensive review of recent progress in evaluating and enhancing multi-turn LLM interactions. Centered on a task-oriented taxonomy-spanning instruction following in domains such as mathematics and coding, and conversational engagement in role-playing, healthcare, education, and adversarial jailbreak settings-we systematically examine the challenges of maintaining context, coherence, fairness, and responsiveness across prolonged dialogues. We organize existing benchmarks and datasets into coherent categories reflecting the evolving landscape of multi-turn dialogue evaluation, and review a broad spectrum of enhancement methodologies, including model-centric strategies (in-context learning, supervised fine-tuning, reinforcement learning, and architectural innovations), external integration approaches (memory augmentation, retrieval-based methods, and knowledge graphs), and agent-based techniques for collaborative interaction. Finally, we identify open challenges and promising directions for future research to further improve the robustness and effectiveness of multi-turn LLM interactions.

cs.CL↗
arXiv2026

Towards Training-Free Scene Text Editing

Scene text editing seeks to modify textual content in natural images while maintaining visual realism and semantic consistency. Existing methods often require task-specific training or paired data, limiting their scalability and adaptability. In this paper, we propose TextFlow, a training-free scene text editing framework that integrates the strengths of Attention Boost (AttnBoost) and Flow Manifold Steering (FMS) to enable flexible, high-fidelity text manipulation without additional training. Specifically, FMS preserves the structural and style consistency by modeling the visual flow of characters and background regions, while AttnBoost enhances the rendering of textual content through attention-based guidance. By jointly leveraging these complementary modules, our approach performs end-to-end text editing through semantic alignment and spatial refinement in a plug-and-play manner. Extensive experiments demonstrate that our framework achieves visual quality and text accuracy comparable to or superior to those of training-based counterparts, generalizing well across diverse scenes and languages. This study advances scene text editing toward a more efficient, generalizable, and training-free paradigm. Code is available at https://github.com/lyb18758/TextFlow

cs.CV↗
arXiv2026

When Documents Disagree: Measuring Institutional Variation in Transplant Guidance with Retrieval-Augmented Language Models

Patient education materials for solid-organ transplantation vary substantially across U.S. centers, yet no systematic method exists to quantify this heterogeneity at scale. We introduce a framework that grounds the same patient questions in different centers' handbooks using retrieval-augmented language models and compares the resulting answers using a five-label consistency taxonomy. Applied to 102 handbooks from 23 centers and 1,115 benchmark questions, the framework quantifies heterogeneity across four dimensions: question, topic, organ, and center. We find that 20.8% of non-absent pairwise comparisons exhibit clinically meaningful divergence, concentrated in condition monitoring and lifestyle topics. Coverage gaps are even more prominent: 96.2% of question-handbook pairs miss relevant content, with reproductive health at 95.1% absence. Center-level divergence profiles are stable and interpretable, where heterogeneity reflects systematic institutional differences, likely due to patient diversity. These findings expose an information gap in transplant patient education materials, with document-grounded medical question answering highlighting opportunities for content improvement.

cs.IR↗
arXiv2026

Consistency of Large Reasoning Models Under Multi-Turn Attacks

Large reasoning models with reasoning capabilities achieve state-of-the-art performance on complex tasks, but their robustness under multi-turn adversarial pressure remains underexplored. We evaluate nine frontier reasoning models under adversarial attacks. Our findings reveal that reasoning confers meaningful but incomplete robustness: most reasoning models studied significantly outperform instruction-tuned baselines, yet all exhibit distinct vulnerability profiles, with misleading suggestions universally effective and social pressure showing model-specific efficacy. Through trajectory analysis, we identify five failure modes (Self-Doubt, Social Conformity, Suggestion Hijacking, Emotional Susceptibility, and Reasoning Fatigue) with the first two accounting for 50% of failures. We further demonstrate that Confidence-Aware Response Generation (CARG), effective for standard LLMs, fails for reasoning models due to overconfidence induced by extended reasoning traces; counterintuitively, random confidence embedding outperforms targeted extraction. Our results highlight that reasoning capabilities do not automatically confer adversarial robustness and that confidence-based defenses require fundamental redesign for reasoning models.

cs.AI↗
arXiv2026

Time-To-Inconsistency: A Survival Analysis of Large Language Model Robustness to Adversarial Attacks

Large Language Models (LLMs) have revolutionized conversational AI, yet their robustness in extended multi-turn dialogues remains poorly understood. Existing evaluation frameworks focus on static benchmarks and single-turn assessments, failing to capture the temporal dynamics of conversational degradation that characterize real-world interactions. In this work, we present a large-scale survival analysis of conversational robustness, modeling failure as a time-to-event process over 36,951 turns from 9 state-of-the-art LLMs on the MT-Consistency benchmark. Our framework combines Cox proportional hazards, Accelerated Failure Time (AFT), and Random Survival Forest models with simple semantic drift features. We find that abrupt prompt-to-prompt semantic drift sharply increases the hazard of inconsistency, whereas cumulative drift is counterintuitively \emph{protective}, suggesting adaptation in conversations that survive multiple shifts. AFT models with model-drift interactions achieve the best combination of discrimination and calibration, and proportional hazards checks reveal systematic violations for key drift covariates, explaining the limitations of Cox-style modeling in this setting. Finally, we show that a lightweight AFT model can be turned into a turn-level risk monitor that flags most failing conversations several turns before the first inconsistent answer while keeping false alerts modest. These results establish survival analysis as a powerful paradigm for evaluating multi-turn robustness and for designing practical safeguards for conversational AI systems.

cs.CL↗