arXiv subjects

Baichuan Li

Publications and source records attributed to Baichuan Li.

15 recordsLinked to original sources

arXiv2026

Auditing Escalation Reliability in Healthcare LLM Agents: A Perturbation-Controlled Benchmark of Workflow-State Failures

Trustworthy biomedical AI requires reliable action selection when evidence is missing, stale, conflicting, or late, not only accurate answers on static inputs. We instantiate an episode-level evaluation protocol as 400 synthetic healthcare workflow episodes: 40 hand-authored semantic seeds across four task classes, two surface parameterizations, and five controlled variants, scored on ANSWER/ASK CLARIFICATION/ABSTAIN/ESCALATE under a pre-declared 5:1 missed-versus-unnecessary escalation cost. Reference actions are single-author study-policy labels, not clinician-adjudicated ground truth. Across five open-weight configurations (1.7B--9.7B), clean decision accuracy reaches 0.97, yet uncertainty-prompted models miss 32--97% of required escalations. The stronger 9.7B configuration narrows but does not remove the gap, falling from 0.97 clean accuracy to 0.78 under conflict. Verbalized confidence is inconsistent (ECE 0.10--0.54; failure AUROC 0.48--0.77), and a development-tuned confidence-only policy yields no utility gain. A transparent workflow-aware controller improves 5:1 utility for every model and reduces missed escalations to 0.12--0.24, at specificity 0.83--0.93. However, a 30-concept lexical stress test reduces each rule detector's recall from 1.00 on trigger phrases to 0.00 on unseen paraphrases, limiting the controller claim to this benchmark. Seed-clustered statistics, ablations, cross-task transfer, raw outputs, and manifests are packaged for audit. The results motivate explicit workflow-state evaluation while showing that apparently effective routing controls need their own generalization tests.

cs.AI↗

arXiv2026

Learning What to Ask and What to Believe: Distilling a Variational Bayesian Agent into Language Models

Interactive tasks such as eliciting a user's preferences or diagnosing a hidden mechanism require an agent to track its uncertainty over a latent state, ask informative questions and act on the answers. Language models have no explicit mechanism for this, whereas Bayesian information seeking provides one through posterior inference and expected information gain (EIG), and we ask whether it can be distilled into an 8B language model. Our teacher, a lightweight Bayesian agent with a sequential variational belief and EIG query selection, matches or approaches exact and particle references on three task families. We generate synthetic interactions with the teacher on episodes disjoint from the benchmark and fine-tune \Student{} to reproduce its queries and its belief state as a structured scratchpad. On three Multi-Turn Puzzles tasks the student raises Circuit Decoding from 0.00 to 0.93 success, lowers the normalized rank of the recommended film from 0.50 to 0.10 and cuts word guessing from 32 to 3.4 attempts, exceeding the strongest reported language models on the two comparable tasks. Exposing the belief state closes much of the gap where the decision is an implicit computation rather than a read-out of the interaction; in matched teacher contrasts, EIG supervision consistently outperforms random querying, whereas greater inference fidelity does not uniformly improve the distilled student; a 70B language-model teacher does not match the Bayesian teacher in downstream student performance, and the learned skill does not transfer to an unseen task family. Variational Bayesian information seeking can thus be amortized into a language model by transferring the belief state that drives its questions and decisions.

cs.LG↗

arXiv2026

Remember, Verify, or Ask? Evaluating Memory Decisions at Reuse Time

Persistent memory can personalize an LLM agent, but an incorrect durable update can silently distort future behavior. We study the memory-clarification boundary: whether interaction-derived information should be persisted, used only in the current context, re-verified, or clarified with the user. MCB contains 140 primary scenarios, split into 70 development and 70 held-out items, plus a separate 70-item contrast set. It evaluates both action labels and structured tool-call selection. Two non-authors independently label the 70 held-out primary and 70 contrast items (97.1% agreement, Cohen's kappa = 0.962); a blind third resolves four disagreements, replacing eight author labels by non-author majority. Across Claude and Qwen, models verify changing facts more reliably than they ask users to resolve ambiguity. Bare Qwen asks on 0/12 clarification items while verifying 12/18 freshness items. Few-shot prompting raises accuracy from 0.557 to 0.771 (paired delta = +0.214, Holm-adjusted exact McNemar p_H = 0.002), yet clarification recall remains 0.333. The policy prompt reduces erroneous persistence from 0.243 to 0.100 (p_H = 0.038), although its accuracy gain is not significant. Label-tool agreement is 57% for each Claude model and 23% for Qwen; Qwen accuracy falls from 0.557 to 0.343 (p_H = 0.047). Memory evaluation must test both stated decisions and tool-call choices.

cs.CL↗

arXiv2026

Execution Realism and Reproducibility in LLM-Based Trading Systems: A Systematic Scoping Review and Evidence Audit

Execution realism remains weakly standardized in research on large-language-model-based trading systems, limiting comparison and reproduction across backtests, simulations, and portfolio benchmarks. This article presents a PRISMA-ScR systematic scoping review with a nested execution-reproducibility evidence audit. Eight reproducible query families returned 796 query-level records; DOI/title deduplication left 687 unique records, 101 reports were sought for retrieval, and 59 full texts were recovered through direct and fallback open-access routes. Fifty-three reports are provisionally eligible for expanded evidence charting, subject to reconciliation of two blind author-validation packets. The charting framework separately evaluates point-in-time controls, temporal splits, held-out evaluation, trading costs and turnover, execution semantics, universe construction, architecture reporting, and artifact availability. Direct-trading, portfolio or alpha-construction, and benchmark studies are retained as distinct descriptive subgroups rather than pooled as a common performance estimand. Search responses, retrieval attempts, file hashes, evidence excerpts, screening decisions, and coding worksheets are archived with the review materials. Final field-level counts are intentionally withheld until author validation is complete.

cs.AI↗

arXiv2026

Automated Prefetching for Object Spatial Programming Using Temporal Trace Graphs

Dependent database accesses can serialize application--database round trips on a request's critical path. Planning across traversal steps requires knowing both which relationships a computation may follow and how discovered objects become subsequent execution contexts. We identify a programming-model contract that exposes this information through inspectable access paths and explicit traversal continuation. We demonstrate this contract in Object-Spatial Programming (OSP), where walkers traverse a typed object graph and execute node-type-specific code. We introduce the Temporal Trace Graph (TTG), a bounded, request-specific graph of candidate objects and traversal-derived transitions. At walker spawn, the TTG Builder composes statically extracted access and continuation rules over the live graph from the request's starting node. An asynchronous prefetcher loads planned objects while the walker executes. We implement TTG in Jac/PostgreSQL and evaluate four applications covering linear, multi-hop, recursive, and value-dependent traversals. TTG improves end-to-end latency in three applications, achieving up to 2.99x speedup over the stock runtime. Comparisons with manually staged prefetching characterize the benefits and costs of automatic cross-step planning. The evaluation also shows that planning overhead can exceed avoided demand latency and that value-dependent control can introduce bounded overfetch.

cs.DB↗

arXiv2026

Engineering Reliable Commit Gates for Agentic AI: Cost-Aware Verification Portfolios under Common-Mode Data Failures

Agentic systems commit state-changing actions, but additional verifiers can inherit the same upstream fault. We present VP-CONTROL, a runtime-assurance design and deterministic benchmark for cost-aware commit gates. Its 48 task templates yield 2,880 scenarios across six fault regimes. A fixed-call 2 x 2 experiment separates verifier-model diversity from evidence-source diversity. On frozen proposals from two local actor families, a cross-model vote over shared evidence approves 62.9% of unsafe proposals, versus 22.9% with an independent source. The source effect is 40.9 percentage points, compared with 11.3 for model diversity. A portfolio controller selects verification plans using only deployment-observable metadata. Approximate cluster-adjusted calibration at a nominal 5% per-task target yields 1.9% unsafe execution and 38.2% automated safe coverage on the locked test. Matched-budget portfolios also improve on fixed verification policies. Transfer remains conditional: unseen fault families yield 16-26% risk, and a FinQA check fails to reproduce the source effect with the tested small verifiers. A preregistered live HTTP/SQLite study tests concurrent writes and lost responses. After-check races defeat verifier-only gates; transactional partial guards prevent only covered failures, while a full atomic guard records no unsafe effects across 216 episodes. Idempotent request identifiers prevent duplicate effects after lost responses. The results motivate explicit evidence lineage, cost-aware selection, and commit-time enforcement, while exposing the limits of approximate calibration and local-tool generalization.

cs.SE↗

arXiv2026

Beyond Helpfulness: A Teaching-over-Solving Diagnostic for Measuring Educational Impact in LLM Tutors

Large language models are increasingly proposed as educational tutors, yet stronger task-solving ability does not necessarily imply stronger learning support. Motivated by recent calls to measure the social impact of NLP systems in practice, we study whether public LLM tutoring benchmarks distinguish learning-supportive behavior from mere answer production. We propose a lightweight diagnostic based on the gap between solving-oriented and pedagogy-oriented benchmark performance. Using public MathTutorBench leaderboard results, we show that these dimensions are only partially aligned: across eight publicly reported models, the correlation between solving and pedagogy composites is 0.421, and several models shift meaningfully in rank when evaluation moves from solving to pedagogy. We then analyze the public TutorBench sample and show that agency-relevant behaviors are explicitly encoded in benchmark rubrics, especially in active-learning settings that reward guiding questions, calibrated hints, and non-disclosive scaffolding. Together, these findings suggest that educational-impact evaluation should not treat task success as a sufficient proxy for learning support. We argue that public tutoring benchmarks can better support positive-impact evaluation by reporting solving-oriented and pedagogy-oriented scores separately and by making disclosure-sensitive, student-agency-preserving criteria more explicit.

cs.AI↗

arXiv2026

Toward Workflow-Aware Benchmarking for Healthcare NLP Agents

Large language model (LLM) agents are increasingly proposed for healthcare tasks such as clinical documentation, evidence retrieval, patient messaging, and care coordination. Yet many evaluations remain limited to static medical question answering or one-shot generation, under-representing longitudinal state, interruptions, and human handoffs. We introduce an episode-level evaluation protocol for healthcare NLP agents. The protocol separates evidence across model, agent, and simulated-workflow behavior; specifies a five-field episode schema; and defines annotation and scoring for state continuity, evidence traceability, and escalation decisions. It is instantiated as four task templates: documentation update, evidence retrieval, patient messaging, and triage handoff. The protocol does not claim to measure clinical outcomes or deployment value. Instead, it supplies a reproducible intermediate evaluation layer between static benchmarks and prospective workflow studies, with an explicit cost-sensitive treatment of missed versus unnecessary escalation.

cs.CL↗

arXiv2026

Reliable Financial Named Entity Recognition Under Domain Shift: Confidence Estimation and Selective Prediction

Financial AI systems often train information extractors on one textual register and deploy them across filings, news, and user-generated content, and standard F1 scores do not indicate which predictions remain safe to automate when that input distribution changes. We study confidence estimation and selective prediction for financial named entity recognition (NER) on a three-tier stress test spanning SEC filings, financial news, and general-topic social media as an extreme out-of-domain condition, evaluating a BERT tagger and LoRA-tuned Qwen2.5-0.5B/1.5B models with five inference-time confidence signals, three training seeds, and bootstrap intervals. Confidence rankings themselves change under shift: whole-output probability is the strongest in-domain error detector but deteriorates out of domain, whereas entity-span probability and self-consistency are more robust; self-consistency is also better calibrated without post-hoc fitting. Abstention reduces sentence error from 34.3% to below 2% on the highest-confidence 40% of in-domain inputs and remains useful on financial news, but recovers no usefully large clean subset under the extreme social-media shift. These results motivate a staged deployment strategy that detects severe distribution shift upstream before applying prediction-level confidence gating.

cs.CL↗

arXiv2026

Stale Does Not Mean Unsafe: Guard Precision for Tool-Using LLM Agents under Infrastructure State Races

Tool-using language-model agents increasingly mutate schedulers, data pipelines, object stores, and access-control systems. Between an agent's read and its commit, external state can change, but not every change makes the commit unsafe. We separate invalidating races, which break a declared safety predicate, from predicate-preserving and irrelevant races, and ask how precisely runtime guards distinguish them. Our deterministic simulator separates visible from authoritative state and injects five non-atomic failure mechanisms across 16 infrastructure tasks in four domains; frozen agent proposals are replayed counterfactually under every controller without an LLM judge. We evaluate three commit-time guard granularities (global epoch, read-set version, semantic commit predicate), multi-level verification, and model-side gates on three locally hosted quantized model families (Qwen3-4B, Phi-4-mini, Gemma4-8B; 3,456 trajectories on one GPU). All three guards eliminate unsafe commits, but their availability differs sharply: freshness-based guards needlessly block 92-95% of benign races, forfeiting up to 43% of safe task completions, while the complete predicate guard blocks none. That precision is contract-dependent: deleting a single declared clause converts exactly its fault family into unsafe commits (up to 7.9%). Model-side signals do not substitute: verbal confidence is miscalibrated (ECE approximately 0.37), action agreement matches a random gate, a cautionary prompt leaves the direct unsafe rate essentially unchanged, and after a freshness-guard block agents re-commit unsafely from refreshed but still-incomplete reads. Under degraded telemetry a hidden concurrent mutation remains observationally clean, bounding every selective policy. Precise runtime enforcement therefore requires semantic contracts, not freshness heuristics or model self-assessment.

cs.AI↗

arXiv2026

Option-Implied Signals and Crash Risk: Predictability and Machine-Learning Evidence from U.S. Equity Options

We re-estimate canonical option-implied predictability evidence using a unified 2015--2026 panel of 12.36 million U.S. equity firm-day observations across 10,026 underlyings. We split the sample into three regimes: late-post-crisis low volatility (2015--2019), high-volatility transition (2020--2022), and AI/mega-cap concentration (2023--2026). The Xing et al. (2010) smirk--return relationship weakens steadily: the next-month univariate smirk coefficient falls from $-0.023$ $(t=-5.5)$ in 2015--2019 to an insignificant $-0.006$ $(t=-1.5)$ in 2023--2026, and turns positive at the three-month horizon $(+0.016,\ t=+2.1)$. In joint specifications with all six canonical signals, the smirk is insignificant throughout and again changes sign in the latest regime. By contrast, the Cremers--Weinbaum IV spread and a Bakshi et al. (2003)-style risk-neutral skewness measure remain significant across regimes and specifications. A boosted-tree benchmark using IV-surface, trading-activity, Greeks, and liquidity features outperforms linear models for next-month return prediction only in the AI/mega-cap regime, with $R^2_{\mathrm{OOS}}=+1.29%$ versus $+0.07%$. Permutation importance identifies different leading predictors by regime, and no canonical hand-engineered signal enters the top five. For firm-level five-day crash classification, AUC is highest in calm markets $(0.706,\ \text{XGBoost})$ and lowest during the high-volatility transition $(0.561)$. Overall, option-implied predictability is real but regime-dependent, and the post-2023 AI/mega-cap period differs sharply from the pre-COVID setting in which the canonical results were established.

q-fin.ST↗

arXiv2026

AutoTour: Automatic Photo Tour Guide with Smartphones and LLMs

We present AutoTour, a system that enhances user exploration by automatically generating fine-grained landmark annotations and descriptive narratives for photos captured by users. The key idea of AutoTour is to fuse visual features extracted from photos with nearby geospatial features queried from open matching databases. Unlike existing tour applications that rely on pre-defined content or proprietary datasets, AutoTour leverages open and extensible data sources to provide scalable and context-aware photo-based guidance. To achieve this, we design a training-free pipeline that first extracts and filters relevant geospatial features around the user's GPS location. It then detects major landmarks in user photos through VLM-based feature detection and projects them into the horizontal spatial plane. A geometric matching algorithm aligns photo features with corresponding geospatial entities based on their estimated distance and direction. The matched features are subsequently grounded and annotated directly on the original photo, accompanied by large language model-generated textual and audio descriptions to provide an informative, tour-like experience. We demonstrate that AutoTour can deliver rich, interpretable annotations for both iconic and lesser-known landmarks, enabling a new form of interactive, context-aware exploration that bridges visual perception and geospatial understanding.

cs.HC↗

arXiv2025

MTP: A Meaning-Typed Language Abstraction for AI-Integrated Programming

Software development is shifting from traditional programming to AI-integrated applications that leverage generative AI and large language models (LLMs) during runtime. However, integrating LLMs remains complex, requiring developers to manually craft prompts and process outputs. Existing tools attempt to assist with prompt engineering, but often introduce additional complexity. This paper presents Meaning-Typed Programming (MTP), a novel paradigm that abstracts LLM integration through intuitive language-level constructs. By leveraging the inherent semantic richness of code, MTP automates prompt generation and response handling without additional developer effort. We introduce the (1) by operator for seamless LLM invocation, (2) MT-IR, a meaning-based intermediate representation for semantic extraction, and (3) MT-Runtime, an automated system for managing LLM interactions. We implement MTP in Jac, a programming language that supersets Python, and find that MTP significantly reduces coding complexity while maintaining accuracy and efficiency. MTP significantly reduces development complexity, lines of code modifications needed, and costs while improving run-time performance and maintaining or exceeding the accuracy of existing approaches. Our user study shows that developers using MTP completed tasks 3.2x faster with 45% fewer lines of code compared to existing frameworks. Moreover, MTP demonstrates resilience even when up to 50% of naming conventions are degraded, demonstrating robustness to suboptimal code. MTP is developed as part of the Jaseci open-source project, and is available under the module byLLM.

cs.PL↗

arXiv2025

It's Not Just Labeling -- A Research on LLM Generated Feedback Interpretability and Image Labeling Sketch Features

The quality of training data is critical to the performance of machine learning applications in domains like transportation, healthcare, and robotics. Accurate image labeling, however, often relies on time-consuming, expert-driven methods with limited feedback. This research introduces a sketch-based annotation approach supported by large language models (LLMs) to reduce technical barriers and enhance accessibility. Using a synthetic dataset, we examine how sketch recognition features relate to LLM feedback metrics, aiming to improve the reliability and interpretability of LLM-assisted labeling. We also explore how prompting strategies and sketch variations influence feedback quality. Our main contribution is a sketch-based virtual assistant that simplifies annotation for non-experts and advances LLM-driven labeling tools in terms of scalability, accessibility, and explainability.

cs.HC↗

arXiv2023

The Jaseci Programming Paradigm and Runtime Stack: Building Scale-out Production Applications Easy and Fast

Today's production scale-out applications include many sub-application components, such as storage backends, logging infrastructure and AI models. These components have drastically different characteristics, are required to work in collaboration, and interface with each other as microservices. This leads to increasingly high complexity in developing, optimizing, configuring, and deploying scale-out applications, raising the barrier to entry for most individuals and small teams. We developed a novel co-designed runtime system, Jaseci, and programming language, Jac, which aims to reduce this complexity. The key design principle throughout Jaseci's design is to raise the level of abstraction by moving as much of the scale-out data management, microservice componentization, and live update complexity into the runtime stack to be automated and optimized automatically. We use real-world AI applications to demonstrate Jaseci's benefit for application performance and developer productivity.

cs.CL↗