Search arXiv⌕ Search

arXiv subjects

Fei Huang

Publications and source records attributed to Fei Huang.

At least 19 recordsLinked to original sources
arXiv2026

From Explaining to Enabling: Actionable Transparency in Algorithmic Insurance Pricing

Machine learning (ML) systems increasingly make consequential decisions about individuals, including credit, hiring, and insurance. Research on ML-driven insurance pricing has asked how the pricing mechanism itself can be made accountable. Yet it remains unclear how an individual policyholder's algorithmically determined premium should be explained to them. We argue that a useful explanation must be valid, actionable, and reliable. A valid recommendation lowers the premium according to the model. An actionable recommendation changes only characteristics the policyholder can control. A reliable recommendation has an actual outcome that is likely to meet the target, even in sparsely observed regions of the feature space where the model is uncertain. We formalize these three requirements as axioms and develop a method that turns an algorithmic premium into recommendations that meet them. The method meets the third requirement with a Conformal Prediction constraint, which screens out recommendations that the model is too uncertain about and thereby bounds the rate of unreliable recommendations. We apply four Counterfactual Explanation techniques to 57,832 real Florida flood-insurance claims and 98,058 synthetic driver-telematics policies. We adapt these techniques for insurance applications, so that their recommendations change only controllable characteristics and resemble policyholders observed in the data. In our illustrations, the constrained recommendations for high-risk policyholders are both more reliable and associated with larger predicted risk reductions than the unconstrained ones. These results suggest that insurers can give policyholders explanations that are both actionable and trustworthy. The work contributes to research on algorithmic accountability and offers implications for insurers and regulators.

cs.LG↗
arXiv2026

ToolOPD: Credit-Aware On-Policy Distillation for Tool-Use Agents

On-policy distillation (OPD) enhances the capabilities of small language agents by providing dense, token-level supervision from a teacher model on student-generated trajectories. However, standard OPD implicitly assumes that credit is uniformly distributed across long-horizon trajectories, assigning the same strength of teacher supervision to different actions. Through credit-distribution analysis and outcome-reversal counterfactual experiments, we find that trajectory credit is concentrated on a small number of critical actions. Moreover, even when the trajectory is held fixed, reversing only the final outcome substantially changes the credit assigned to multiple interaction turns. These findings show that the uniform-credit assumption fails to identify critical decisions and ultimately degrades the model's reasoning accuracy. To address these challenges, we propose ToolOPD, an outcome-aware on-policy distillation framework designed for long-horizon tool use. ToolOPD first summarizes each completed trajectory and its outcome into a reusable hindsight skill. It then measures how this skill changes the teacher's probability of each student action token and converts the change into outcome-aware credit. The resulting credit preserves supervision that supports successful decisions or discourages failed ones, while suppressing irrelevant and conflicting signals. ToolOPD uses these credits to reweight the action-level distillation loss and jointly optimizes it with reinforcement learning. Extensive experiments across multiple long-horizon tool interaction benchmarks demonstrate that ToolOPD significantly outperforms existing reinforcement learning and self-distillation baselines.

cs.AI↗
arXiv2026

Gauge Interactions Involving New Vector and Pseudoscalar Particles

The Vector gauge bosons $Z'$ and axion-like particles (ALPs) $a$ are predicted by the $U(1)_X$ gauge extension and Peccei-Quinn (PQ) models, respectively. We focus on the interactions of the $Z'$ boson and the $a$ particle with the $W^\pm$ gauge boson, and investigate the production cross section and related kinematic distributions for the $Z'/a\to W^+W^-$ decay process at hadron colliders. We consider the angular distributions of charged leptons in the final states corresponding to the semileptonic and fully leptonic decays of the $W^\pm$ boson, separately. The study shows that, in the associated production processes of $Z'/a$ with $W^+$ or $Z$ bosons, the angular distributions of final-state leptons can serve as an important observable for probing new gauge interactions, and can also provide a criterion to distinguish vector particles from pseudoscalar particles.

hep-ph↗
arXiv2026

Gravitational Waves as a Probe of Multi-Fluid Cosmologies: General Formalism and the Case of Cosmological Stasis

Scalar-induced gravitational waves (SIGWs) provide a powerful probe of the expansion history of the universe prior to the Big Bang. Indeed, the production of such gravity waves depends on the evolution of first-order scalar perturbations within the early universe, while their subsequent propagation depends on the background equation of state. In this paper we extend previous discussions by studying SIGWs in cosmologies containing multiple fluids. As we shall demonstrate, the gravitational-wave signal within such multi-fluid cosmologies is modified in two distinct ways: not only are there new velocity effects which have no single-fluid analogues, but there are also significant alterations to the subsequent redshifting of modes that enter the horizon before the multi-fluid era ends. Thus SIGWs provide a probe of not only the overall expansion rate within such cosmologies, but also their internal multi-fluid composition. As an example, we then consider the case of gravitational waves arising during an epoch of cosmological stasis. During such an epoch, the abundances of energy components with different equations of state remain constant, thereby giving rise to a universe with a constant effective equation of state which nevertheless remains genuinely multi-fluid despite cosmological expansion. Despite their constant equations of state, we find that such stasis cosmologies also give rise to tensor power spectra and present-day gravitational-wave abundances which are significantly different from those expected within a standard radiation-dominated universe.

astro-ph.CO↗
arXiv2026

$Z^\prime$ Portal Dark Matter with Observable $ΔN_{\rm eff}$

In the conventional $Z^\prime$ portal dark matter scenario, the prediction of detectable dark matter $χ$ typically relies on the collider sensitivities of $Z^\prime$ and direct detection, where the Majorana type right-handed neutrinos are usually assumed. However, if the right-handed neutrinos $ν_R$ are Dirac type, they will contribute to the additional effective number of relativistic species $ΔN_{\rm eff}$, which brings different detectable predictions for $Z^\prime$ portal dark matter. In light of the great improvement of $ΔN_{\rm eff}$ for the upcoming experiments, we investigate the $Z^\prime$ portal dark matter with Dirac type $ν_R$. Under the $U(1)_{B-L}$ symmetry, this model includes $ν_R$ with $U(1)_{B-L}$ charge $Q_{ν_R}=-1$ and $χ$ with arbitrary $Q_χ$ beyond the SM. Based on the relation in the production of $χ$ and $ν_R$, both the WIMP and FIMP dark matter through the $Z^\prime$ portal scenario are considered. We perform a comprehensive exploration of the viable parameter space under the constraints from $ΔN_{\rm eff}$ induced by thermal and non-thermal $ν_R$, perturbative limit, dark matter direct and indirect detection, and collider searches of $Z^\prime$.

hep-ph↗
arXiv2026

Model What Matters: Decision-Guided Selective World Modeling for Language Agents

Reinforcement learning provides a direct mechanism for optimizing task objectives in large language model-driven agents, but task rewards are typically sparse trajectory-level signals that only evaluate task success, without directly supervising action-conditioned environment transitions, leaving the model with limited explicit understanding of action consequences and environment dynamics. Recent work introduces auxiliary supervision through policy and world modeling co-training, yet typically treats complete environmental observations uniformly as world modeling targets, overlooking the varying decision relevance of observation content. We propose DGS-WM (Decision-Guided Selective World Modeling), a decision-guided policy-world modeling co-training approach for language agents. DGS-WM employs a Decision-Guided Observation Selection component that filters observation units with decision value as world modeling targets based on their support for subsequent policy responses, and a Confidence-Adaptive Weighting component that adaptively adjusts supervision strength according to the model's prediction confidence on selected targets. Together, these components form a decision-aware auxiliary world modeling loss, jointly optimized with the RL loss on the same policy model. Across multiple benchmarks covering single-hop/multi-hop question answering and interactive web tasks, DGS-WM achieves consistent improvements under different model scales and RL algorithms, demonstrating more stable and effective performance compared to full-observation world modeling methods.

cs.AI↗
arXiv2026

Spotting Stasis in Cosmological Perturbations

As discussed in a number of recent papers, cosmological stasis is a phenomenon wherein the abundances of multiple cosmological energy components with different equations of state remain constant for an extended period despite the expansion of the universe. One of the most intriguing aspects of the stasis phenomenon is that it can give rise to cosmological epochs in which the effective equation-of-state parameter $\langle w \rangle$ for the universe is constant, but differs from the canonical values associated with matter, radiation, vacuum energy, etc. Indeed, during such a stasis epoch, the spatial average of the energy density of the universe evolves in precisely the same manner as it would have evolved if the universe were instead dominated by a perfect fluid with an equation-of-state parameter equal to $\langle w \rangle$. However, as we shall demonstrate, this equivalence is broken at the level of the perturbations of the energy density. To illustrate this point, we consider a stasis epoch involving matter and radiation and demonstrate that within this stasis background the density perturbations associated with a spectator matter component with exceedingly small energy density exhibit a power-law growth that persists across the entire duration of the stasis epoch. This growth can potentially lead to significant enhancements of structure at small scales. Such enhancements are not only interesting in their own right, but may also provide a way of observationally distinguishing between a stasis epoch and an epoch of perfect-fluid domination -- even if the universe has the same equation of state in both cases.

astro-ph.CO↗
arXiv2026

Mind the Spatial Gap: A Benchmark for Geospatial Reasoning and Constrained Spatial Editing

Understanding and manipulating spatial relations is fundamental to AI systems that reason about physical environments and construct grounded world models. However, existing evaluations of multimodal large language models (MLLMs) for spatial reasoning rely predominantly on static question-answering (QA) over small-scale or synthetic scenes, rarely probing whether models can modify spatial configurations while satisfying explicit geometric constraints. We introduce GeoSpatialBench, a benchmark designed to evaluate an escalating progression of spatial cognition, ranging from spatial relation understanding and geometric reasoning to constrained spatial editing. Grounded in real-world urban data across six geographically diverse cities, GeoSpatialBench evaluates a comprehensive taxonomy of 28 topological, directional, metric, and ordering relations across structured GeoJSON and map image modalities, with ground-truth references and constraints verified via GIS operations. Our empirical evaluation across frontier MLLMs uncovers a stark spatial gap: structured GeoJSON consistently outperforms map images, with multimodal combinations offering no reliable gain over GeoJSON alone, and while models show reasonable competence on qualitative topological and ordering, performance drops on quantitative metric relations. Crucially, editing performance on ordering relations declines relative to QA, and most failed edits violate at least one specified spatial constraint. These results highlight a gap between passively recognizing spatial relations and actively manipulating geometric configurations. GeoSpatialBench provides an open, rigorous diagnostic testbed to benchmark these spatial limitations, laying essential empirical groundwork for physically grounded world models and embodied spatial AI.

cs.AI↗
arXiv2026

Deeper is Not Always Better: Mitigating the Alignment Tax via Confident Layer Decoding

Autoregressive generation in large language models (LLMs) conventionally decodes from the final layer, assuming that deeper representations yield more reliable next-token predictions. We revisit this assumption by revealing a recurring Guess-Refine-Perturb dynamic: early layers form coarse guesses, intermediate layers refine reasoning-relevant semantics, and final layers can perturb these refined predictions toward generic or alignment-preferred tokens. We introduce Confident Decoding, a training-free decoding strategy that dynamically selects the most reliable near-final layer through entropy-guided conservative backward search. We further provide a theoretical formulation of layer selection as an optimal stopping problem, showing that under bounded projection noise and dominant late-stage alignment perturbation, our search rule filters perturbation while bounding the loss relative to the oracle refinement layer. Experiments across dense and Mixture-of-Experts LLMs demonstrate consistent gains on challenging reasoning benchmarks, including GPQA-Diamond, Omni-MATH, and HLE, with zero memory overhead and less than 2% latency increase. These results suggest dynamically bypassing final-layer perturbations can unlock stronger reasoning behavior from aligned LLMs.

cs.CL↗
arXiv2026

MATCH: Model-Aware Tool Learning with Curriculum Scheduling and Hierarchically Gated Rewards

Tool learning enables large language models (LLMs) to use external tools for tasks beyond parametric knowledge. Reinforcement learning can optimize tool-call behavior from feedback, but current methods still face two problems: fixed-threshold curricula can become misaligned with the policy's evolving capability boundary, and additive rewards can leak argument-level credit when the predicted tool is wrong. To address these problems, we propose MATCH, a closed-loop framework for model-aware tool learning with curriculum scheduling and hierarchically gated rewards. Model-Aware Curriculum Learning (MACL) maintains reward-derived sample difficulty that co-evolves with the policy, and each epoch selects samples near the current capability boundary together with a top-k pool of harder cases. Hierarchical Tool-call Gated Reward (HTGR) scores tool name, argument key, and argument value as a gated chain, granting credit at each level only when prerequisites hold. The same HTGR rewards drive both GRPO updates and MACL's difficulty refresh, closing the loop between policy optimization and sample scheduling. On API-Bank and BFCL V3, MATCH reaches 72.19% and 62.87% overall accuracy, outperforming the main supervised and RL-based baselines. Backbone experiments further show consistent improvements across four backbones from two model families.

cs.LG↗
arXiv2026

Think Thrice Before Reranking: Multi-perspective Evidence and Reasoning Integration for Text Reranking

Reasoning-based reranking with Large Language Models (LLMs) has shown promising improvements in text ranking. However, current methods predominantly rely on a single reasoning trajectory, resulting in rankings that are susceptible to reasoning errors and inherently constrained in modeling the multifaceted signals underlying document relevance. To resolve this dilemma, we propose MERIT-Rank(Multi-perspective Evidence and Reasoning Integration for Text Reranking), a framework that models complementary reasoning trajectories to improve reranking robustness. MERIT-Rank formulates a Multi-Trajectory Reasoning Space (MTRS) that evaluates query-document relevance from multiple perspectives and introduces a joint reranker that consolidates these reasoning paths into a unified ranking decision. We further develop Progressive Rank Policy Optimization (PRPO), a progressive training framework that stabilizes reasoning trajectories while continually improving ranking quality through staged optimization objectives. Experiments on both reasoning-intensive and traditional retrieval benchmarks show that MERIT-Rank consistently achieves superior performance over competitive baselines. The 4B model notably outperforms most 7B and even 32B rerankers on BRIGHT.

cs.IR↗
arXiv2026

The Imitation Game: When LLMs Learn to Reason Like Programs via Code-Centric Reasoning Data Synthesis

Large Language Models (LLMs) excel at programming tasks but frequently fail at deterministic, fine-grained reasoning in natural language, relying heavily on semantic approximations rather than robust symbolic execution. To bridge this gap, we propose MIMIC, a framework that leverages executable code as a rigorous medium for reasoning data synthesis. MIMIC fundamentally transforms algorithms into verifiable reasoning trajectories through narrative fusion, code-guided test synthesis, and dynamic code instrumentation. Crucially, these explicit intermediate execution states naturally form a Code-Instrumented Reward (CIR), providing dense, high-fidelity process supervision for reinforcement learning without external reward models. Extensive evaluations reveal that models trained via SFT and GRPO on our synthesized dataset achieve substantial, consistent gains. Our method significantly elevates accuracy across general reasoning, complex mathematical benchmarks, and fine-grained deterministic tasks, demonstrating that the procedural rigor of executable code can effectively unlock and enhance the generalized reasoning capabilities of LLMs. Our code and data are available at https://github.com/zjy1298/MIMIC.

cs.CL↗
arXiv2026

LightNav-0: Eliciting VLM Spatial Intelligence for Generalist Embodied Navigation

Embodied navigation requires agents to translate heterogeneous goals and visual observations into actions across tasks, environments, and robot embodiments. Modern vision-language models (VLMs) already encode spatial priors for visual grounding, spatial reasoning, and pointing, but these capabilities are rarely elicited directly for robot control. Existing navigation systems instead rely on task- or embodiment-specific components, fragmenting perception, reasoning, and action while offering limited generalization. Here we present LightNav-0, a compact generalist embodied navigation model that elicits the spatial intelligence of a pretrained VLM and aligns it with navigation, without task-specific prediction heads. LightNav-0 represents diverse navigation tasks through a unified token interface: dual-channel pointing expresses task-, scene-, and embodiment-agnostic spatial intent, while a residual vector-quantized action tokenizer maps this intent to precise, embodiment-specific trajectories. Together with temporally aware visual history compression, ER mid-training, supervised fine-tuning, and reinforcement learning, this formulation supports instruction following, open-vocabulary object navigation, and visual tracking within a single model. The navigation training corpus spans 2K+ scenes and 4K+ hours of embodied navigation data. LightNav-ER, the embodied-reasoning checkpoint used to initialize LightNav-0, attains the highest complete-set average across 8 embodied-reasoning benchmarks, while LightNav-0 achieves state-of-the-art monocular success rates across all 10 public navigation simulation settings. Real-world evaluations further demonstrate zero-shot generalization across robot embodiments, diverse scenes, and static and dynamic targets. These results establish compact VLMs as a unified and transferable backbone for generalist embodied navigation.

cs.RO↗
arXiv2026

Terminal-Universe: Turning Agent Trajectories into Scalable Terminal Environments

As terminal-based code agents become prevalent, agent trajectories have accumulated at scale, while realistic, executable environments remain scarce. However, environments are what agent post-training actually requires: each can be re-queried into many verifiable tasks and provides execution feedback, whereas a trajectory is a single frozen demonstration. Rather than generating environments from scratch, we observe that the tool-execution history in existing trajectories exposes the structure and contents of the environments in which they ran, making it possible to reconstruct those environments from the trajectories themselves. Thus, we introduce Terminal-Universe, a framework which turns each trajectory into a reusable environment and explores it for synthesizing new tasks and continued interactions. Specifically, Terminal-Universe replays the file operations recorded in a trajectory to restore each file before the agent modified it, yielding a partial workspace; a completion agent then supplies the missing files and dependencies. On this recovered workspace, we both reconstruct the original intent task and synthesize entirely new ones. Besides, we also scale the tasks along two complementary axes: breadth and depth. For breadth, we mine directional dependency relations between related environments and synthesize cross-workspace queries spanning multiple codebases, as developers routinely do in real-world development. For depth, we extend the initial single-turn query into a multi-round session that captures iterative user feedback and requirement refinement via a user agent. Applied to public terminal agent trajectories, Terminal-Universe produces 37.3k task-sufficient environments. Supervised fine-tuning of Qwen3.5-27B on this corpus improves single-round performance on Terminal-Bench 2.1 by 11.9 points and multi-round performance on EvoCode-Bench v2 MT@4 by 13.8 points.

cs.AI↗
arXiv2026

LLP: LLM-Based Product Pricing in E-commerce

Unlike Business-to-Consumer e-commerce platforms (e.g., Amazon), inexperienced individual sellers on Consumer-to-Consumer platforms (e.g., eBay) often face significant challenges in setting prices for their second-hand products efficiently. Therefore, numerous studies have been proposed for automating price prediction. However, most of them are based on static regression models, which suffer from poor generalization performance and fail to capture market dynamics (e.g., the price of a used iPhone decreases over time). Inspired by recent breakthroughs in Large Language Models (LLMs), we introduce LLP, the first LLM-based generative framework for second-hand product pricing. LLP first retrieves similar products to better align with the dynamic market change. Afterwards, it leverages the LLMs' nuanced understanding of key pricing information in free-form text to generate accurate price suggestions. To strengthen the LLMs' domain reasoning over retrieved products, we apply a two-stage optimization, supervised fine-tuning (SFT) followed by group relative policy optimization (GRPO), on a dataset built via bidirectional reasoning. Moreover, LLP employs a confidence-based filtering mechanism to reject unreliable price suggestions. Extensive experiments demonstrate that LLP substantially surpasses existing methods while generalizing well to unseen categories. We have successfully deployed LLP on Xianyu\footnote\{Xianyu is China's largest second-hand e-commerce platform.\}, significantly outperforming the previous pricing method. Under the same 30\% product coverage, it raises the static adoption rate (SAR) from 40\% to 72\%, and maintains a strong SAR of 47\% even at 90\% recall.

cs.CL↗
arXiv2026

CateKV: On Sequential Consistency for Long-Context LLM Inference Acceleration

Large language models (LLMs) have demonstrated strong capabilities in handling long-context tasks, but processing such long contexts remains challenging due to the substantial memory requirements and inference latency. In this work, we discover that certain attention heads exhibit sequential consistency in their attention patterns, which can be persistently identified using a coefficient-of-variation-based algorithm. Inspired by this observation, we propose CateKV, a hybrid KV cache method that retains only critical token information for consistent heads, thereby reducing KV cache size and computational overhead, while preserving the majority of KV pairs in adaptive heads to ensure high accuracy. We show the unique characteristics of our algorithm and its extension with existing acceleration methods. Comprehensive evaluations on long-context benchmarks show that, while maintaining accuracy comparable to full attention, CateKV reduces memory usage by up to $2.72\times$ and accelerates decoding by $2.18\times$ in single-sample inputs, and boosts throughput by $3.96\times$ in batch scenarios.

cs.LG↗
arXiv2026

SkillNet: Create, Evaluate, and Connect AI Skills

Current AI agents can flexibly invoke tools and execute complex tasks, yet their long-term advancement is hindered by the lack of systematic accumulation and transfer of skills. Without a unified mechanism for skill consolidation, agents frequently ``reinvent the wheel'', rediscovering solutions in isolated contexts without leveraging prior strategies. To address this challenge, we introduce SkillNet, an open infrastructure for creating, evaluating, and organizing AI skills at scale. SkillNet structures skills within a unified ontology that supports creating skills from heterogeneous sources, establishing rich relational connections, and performing multi-dimensional evaluation across Safety, Completeness, Executability, Maintainability, and Cost-awareness. Our infrastructure integrates a repository of over 600,000 skills, an interactive platform, and a versatile Python toolkit. Experiments on ALFWorld, WebShop, and ScienceWorld show 40% higher average rewards and 30% fewer execution steps across multiple backbone models. Furthermore, SkillNet-Gym benchmarks skill retrieval, utilization, and composition, while SkillNet-Fabric enables task-specific skill routing through lightweight Wikis. By formalizing skills as evolving, composable assets, SkillNet provides a robust foundation for agents to move from transient experience to durable mastery.

cs.AI↗
arXiv2026

Coalition-Aware Skill Reliability for Self-Evolving Agents

Agent skills, structured artifacts distilled from interaction trajectories and dynamically reused from skill banks, have become a central mechanism for enabling large language model (LLM)-based self-evolving agents to learn from past experience. Yet existing work has largely focused on the operational aspects of skills, such as acquisition, evolution, and retrieval, while leaving a more fundamental reliability question unresolved: Do accumulated skills in an agent's skill bank actually make positive mechanistic contributions? We investigate this question through systematic skill-bank audits across alternative bank compositions and deployment domains, measuring the resulting changes in agent behavior. These audits reveal two recurring reliability failures: coalition pollution, where bank-level gains conceal negative coalition-level skill contributions, and cross-domain utility reversal, where source-beneficial skills reverse their effects after transfer. These findings motivate two reliability interventions: coalition-aware skill selection during skill accumulation and label-free skill masking after transfer. Coalition-Aware Skill Selection (CASS) selects more reliable candidate skills for the current bank using sampled Shapley marginals. Unsupervised Skill-Masked Coalition Optimizer (u-SMCO) masks transferred skills whose exclusion improves retrieval quality on unlabeled target-domain data. Agentic experiments on LoCoMo, LongMemEval, HotpotQA, and ALFWorld show that CASS and u-SMCO consistently improve task performance and cross-domain generalization over strong skill-based self-evolving agent baselines. Beyond accuracy, coalition-conditioned reliability modeling reduces sensitivity to noisy outcome-reward fluctuations during reinforcement learning and exposes the limits of isolation-based skill evaluation.

cs.AI↗