Searcharxiv⌕ Search

arXiv subjects

Chenxu Wang

Publications and source records attributed to Chenxu Wang.

At least 19 recordsLinked to original sources

PACD: Provenance-Aware Consensus Decoding for Mitigating Cross-Modal Hallucinations in Omnimodal Large Language Models

Omnimodal large language models (Omni-LLMs) integrate audio, video, and text for multimodal reasoning, yet remain vulnerable to cross-modal hallucinations, where one modality improperly influences predictions about another. Existing training-free decoding methods mitigate these errors through modality perturbation or relevance-based weighting. However, due to interference from irrelevant modality cues during fusion and difficulty detecting shared errors, these methods may suppress correct unimodal predictions and allow erroneous consensus to persist. To address these limitations, we propose Provenance-Aware Consensus Decoding (PACD), a training-free framework for mitigating cross-modal hallucinations. PACD comprises two complementary paths based on whether the branches reach consensus. For non-consensus cases, Relevance-Guided Candidate Arbitration identifies task-relevant unimodal evidence and selectively arbitrates between predictions using alternative candidates. For consensus cases, Provenance-Based Consensus Recovery leverages the outranking principle to evaluate the consistency and strength of layer-wise modality evidence for alternative answers, selectively resolving erroneous consensus under confidence constraints. Extensive experiments on four Omni-LLMs and five benchmarks across two tasks demonstrate the superiority of PACD. Compared to existing methods, PACD achieves up to 27.6% relative improvement over standard decoding across CMM, AVHBench, and SVHalluc for cross-modal hallucination mitigation, and an average relative improvement of 7.7% over best baseline across OmniVideoBench and FutureOmni for general audio-visual reasoning.

cs.AI↗

A Comprehensive Survey of Agent Distillation across Models, Artifacts, and Harnesses

Knowledge distillation provides a teacher-student framework for transferring capabilities at reduced cost and has been widely studied for large language models (LLMs). As LLM-based agents move from response generation to sustained, closed-loop task solving, distillation expands beyond model parameters to knowledge retained in reusable resources and execution workflows. This shift from model-level to agent-level knowledge transfer supports efficient deployment, broader capability access, and continual self-evolution. However, existing surveys largely examine model distillation and agent system design separately, limiting comparison of transfer strategies across agent components. To bridge this gap, we formalize Agent Distillation as persistent knowledge transfer between teacher and student agents and present its first unified survey. Distinguishing experience used for transfer from knowledge retained for reuse, we develop a knowledge-substrate taxonomy based on where knowledge persists: models, artifacts, or harnesses. Building on this taxonomy, we analyze cross-substrate transfer paths and hybrid strategies. To assess the resulting capabilities, we review benchmarks and evaluation protocols for retention, causal attribution, and deployed utility without further teacher support. This survey provides a unified basis for studying agent knowledge transfer and developing efficient, adaptable, and self-evolving agents.

cs.AI↗

DualEvolve: Dual-Level Self-Evolution of Mechanism and Orchestration for Recommendation

Recommender systems are widely used in e-commerce and content platforms, where improvement requires repeated model design, training optimization, and experimental validation. Large language model (LLM) agents increasingly automate this development cycle. Sustained autonomous improvement requires agents to improve both recommendation models and their development workflows. However, task-performance-driven approaches select updates mainly by candidates' final scores, without a separate optimization signal for experimentation quality. We therefore present DualEvolve, a dual-level framework that co-evolves recommendation mechanisms and agent orchestration toward recursive self-improvement. Inner evolution revises model and training code using task feedback to improve recommendation performance. Outer evolution scores the inner-loop experimental process, updates harness and skill rules using these scores and execution traces, and adjusts orchestration choices based on next-round scores to improve the quality of subsequent experiments. Experiments on MovieLens-1M and three Amazon Reviews 2023 subsets cover four recommendation methods and 16 settings. DualEvolve achieves higher test means than Base, OpenEvolve, and DeepEvolve in all 64 setting-metric comparisons, with macro-averaged relative gains over DeepEvolve of 7.6% (HR@5) and 9.0% (NDCG@5).

cs.IR↗

SPADE: Spatiotemporal Preference Alignment and Diversity Enhancement for Local Search

Proactive query recommendation aims to suggest relevant search queries to users on the homepage of local service apps. Unlike conventional recommendation over a fixed item set, this requires open-set generation from a vocabulary of millions, where outputs must simultaneously align with user preference, match fine-grained spatiotemporal context, cover long-tail intents, and provide diverse alternatives within a few candidate slots. Existing approaches address these axes in isolation: standard tokenization suppresses long-tail queries, coarse location encoding misses fine-grained geographic demand, generic preference optimization fails to align with production impression-click distributions, and standard beam search collapses to near-duplicate outputs. We present SPADE (Spatiotemporal Preference Alignment and Diversity Enhancement), an end-to-end generative framework that jointly optimizes the generation space, contextual representation, preference distribution, and output slate quality through four components: Objective-Aligned BPE (OA-BPE) aligns subword merging with business value signals to boost long-tail query generation probability; Spatio-Temporal Adaptive Representation (STAR) replaces coarse location tokens with continuous multi-scale encodings and soft gating for precise spatiotemporal grounding; MARS-DPO (Multi-dimensional Adaptive Reward Sampling DPO) combines cross-domain pre-training with preference alignment on production impression-click data via composite rewards and adaptive hard negative sampling; and Sigmoid Diversity Beam Search (SDBS) improves intent coverage without sacrificing recall. Offline experiments on large-scale production data show that SPADE outperforms the strongest generative baseline by +5.6% HR@5, -40.3% GeoC@5, and +1.9% Exec@5. A one-week online A/B test confirms production impact: +1.88% Module CTR, +0.177% Search Volume, and +0.105% Search DAU.

cs.IR↗

CityLift: Compositional 3D Urban Generation from a Single Image

Generating editable 3D urban scenes from a single image requires detailed building assets with a coherent spatial arrangement. Occlusion and perspective ambiguity make it difficult to infer complete building geometry together with consistent relative scales and positions. We present CityLift, a framework that generates individual textured building meshes from an urban image by coupling scene-level spatial modeling with building-level synthesis. CityLift first learns building shape and placement jointly through coupled scene and instance streams, producing coarse sparse structures in a common coordinate system. We then introduce generative spatial inheritance, which maps each building's coarse occupancy and scene-conditioned 3D feature field into a local volume to guide a flow model that generates finer structure. The same coordinate mapping places the generated meshes back into the scene, preserving their spatial relationship to the scaffold. To address the scarcity of urban data pairing scene images with individual 3D buildings, we construct CityLift-5K, comprising 4,997 synthetic scenes with instance masks, building meshes, and placement annotations. Experiments on CityLift-5K and scene blocks from UrbanScene3D and InstanceBuilding demonstrate improved building geometry and scene composition. The project page is available at https://ranhaogong.github.io/CityLift/.

cs.CV↗

ReVi: Recursive Recap Video Generation

Novels tell rich stories, but their textual form cannot directly present the sights and sounds of the worlds they describe. Film adaptations offer an audiovisual experience but require substantial production effort and viewing time; film recaps are shorter but depend on existing footage. Neither route directly provides a concise audiovisual account of a novel. We therefore study novel-to-recap video generation: producing a self-contained, source-grounded audiovisual retelling directly from a novel within a fixed viewing-time budget. This task requires compressing the story while preserving causal connections, allocating sufficient time to retained events, and aligning narration with newly generated visuals. We propose ReVi, a dual-loop multi-agent framework guided by edit first, generate second. Its Inner Loop plans story content, narration time, and corresponding shots before video generation. Across novels, its Outer Loop performs recursive self-improvement by revising both the editing policy and the meta policy that governs its revision. Experiments on two evaluation sets show that ReVi improves recap quality and duration compliance over representative baselines, with human ratings and multi-round improvement results providing further support.

cs.MM↗

Risk-aware Road-network Trajectory Synthesis under Local Differential Privacy

Road-network trajectory synthesis under local differential privacy (LDP) enables mobility data analysis without uploading raw trajectories to an untrusted server. Existing locally private trajectory synthesis methods mainly model whole trajectories, point sequences, or grid-level mobility patterns. However, they cannot handle non-uniform privacy risks caused by local road-network and trajectory contexts, such as endpoints, long-stay-associated regions, and low-out-degree branches. They also do not directly protect conditional next-hop choices on road networks, where a user's selected successor may reveal sensitive local movement intentions. To address these challenges, we propose TrajRACE, a locally private synthesis framework for road-network trajectories. TrajRACE uses start, segment-count, and transition reports to represent trajectories. For each transition, it keeps the current road segment as a public aggregation context and perturbs the true successor among legal successors on the client side. It then assigns risk-aware protection strengths to transitions according to local road-network and trajectory contexts. Finally, the server recovers population-level start, segment-count, and conditional transition statistics from heterogeneous privatized reports, and generates synthetic trajectories that follow the public road-network topology. Experiments on three road-network trajectory datasets demonstrate the effectiveness of TrajRACE in improving synthetic-data utility across distributional, query-based, and pattern-level metrics, while providing stronger protection for high-risk local transitions.

cs.CR↗

WinARDet: Compressed Window Auto-Regression for Long-Horizon Vision-Language Grounding

MLLM-based detectors formulate vision-language grounding as autoregressive coordinate generation, yet in dense scenes, they often collapse into a box repetition loop, repeatedly emitting identical boxes until the decoding budget is exhausted. We trace this failure to historical-answer attention takeover: Under full-history autoregression, repeated coordinate tokens in the growing answer history attract disproportionate attention, crowding out the attention sinks needed to attend to new visual evidence. % To break this loop, in this paper, we propose \textbf{Compressed Window Auto-Regression} (WinAR). This new mechanism partitions a long answer sequence into consecutive generation windows and restricts full token-level autoregression to the current window only while compressing each previous window into a fixed set of latent tokens that condition subsequent decoding. This design reduces the repeated exposure to raw historical coordinates that triggers the attention takeover, yet preserves cross-window dependencies essential for coherent multi-object generation. On RSDense, which contains 200-2,000 objects per image, our model WinARDet improves RexOmni F1-score (IoU=0.5) from 32.16 to 39.36 over the baseline, while reducing the redundancy rate from 35.48% to 13.00%. WinARDet also generalizes across decoding strategies and output formats, including parallel box decoding and polygon generation, providing a practical paradigm for long-horizon vision-language grounding across scene densities.

cs.CV↗

AgentSafeEvo: Adversarial Co-Evolution for LLM Agent Safety

Large language model (LLM) agents accomplish complex tasks through multi-turn interactions with external tools. However, malicious content embedded in user queries, tool descriptions, or environment observations can manipulate their decisions and induce unsafe actions. Adversarial co-evolution has emerged as an effective approach to strengthening LLM safety, but existing work primarily focuses on language-level safety or indirect prompt injection in web tasks. In this paper, we introduce AgentSafeEvo, a co-evolutionary reinforcement learning framework for agent safety built on an attacker-defender game spanning diverse tasks and attack types. The attacker learns to construct adversarial agent tasks that expose weaknesses in the defender's tool use, while the defender learns to complete tasks safely under these attacks. We further introduce learnability-guided task weighting, which uses within-task defender-reward variability to adjust each task's contribution to attacker and defender updates. Extensive experiments on agent safety and general capability benchmarks demonstrate that our method improves adversarial robustness while maintaining general agent capabilities.

cs.AI↗

LintEval: Benchmarking LLMs on Linter-Guided File Refactoring

Linters are widely used in code maintenance and real-time IDE editing to surface lint alerts related to performance, style, error-prone, and other patterns in source files. However, existing code benchmarks do not explicitly evaluate linter-guided file refactoring, where models must reduce lint alerts in source files with correct initial functionality while satisfying functionality preservation. We introduce LintEval, a multilingual benchmark for linter-guided file refactoring across C++, Python, and Java, comprising 13,061 programs with correct initial functionality and 171,662 lint alerts extracted using language-specific linters. LintEval jointly evaluates functionality preservation, lint alert reduction, new alert avoidance, and edit minimality. Experiments reveal a persistent trade-off: closed-source commercial models reach up to 90.4% Reduction@P but achieve only 67.1%-68.3% Pass@1, while several open-source small models show low or negative Reduction@P. A controlled fine-tuning study further shows that supervision format matters: refactoring-oriented SFT raises Qwen2.5-Coder-7B Reduction@P from 24.7% to 76.0%. LintEval enables controlled evaluation of models that improve source files under functionality preservation.

cs.SE↗

Information-Gain Calibrated On-Policy Distillation for Deep Search Agents

On-policy distillation (OPD) has recently emerged as a promising approach for transferring the capabilities of large agents to compact students. In deep search, where agents iteratively retrieve external evidence, a teacher's preference among plausible queries before retrieval may not predict which queries will return useful evidence. Standard OPD nevertheless supervises each student query from the preceding context alone, inducing a failure mode we term \textbf{\textit{evidence-blind supervision (EBS)}}. To understand this, we regroup the OPD objective by interaction turns and reveal two implicit priors: a \textit{budget prior} that allocates supervision by token volume, and a \textit{direction prior} that favors tokens preferred by the teacher, neither of which accounts for the evidence returned by the current query. Our empirical analysis shows that both priors contribute to EBS in practice: the first search turn accounts for 51.1\% of total absolute information gain but receives only 18.0\% of the supervision budget, while 43.7\% of search queries have teacher rewards that disagree in sign with post-retrieval information utility. To mitigate this, we propose \ours, which uses post-retrieval information gain (IG) to calibrate both priors: reallocating supervision toward information-critical turns while calibrating query preferences using the utility of retrieved evidence. Experiments on six multimodal search benchmarks show that \ours improves average accuracy by 3.4 and 1.7 percentage points for 2B and 4B students over standard OPD, enabling the 4B student to surpass its 8B teacher on MMSearch and BrowseComp-VL.

cs.IR↗

When Updating Stops Being Learning: Rethinking LLM Self-Evolution via learnable information gain

Self-evolution lets large language models (LLMs) improve iteratively using their own generated data, but often suffers from self-evolution degeneration: performance improves, plateaus, then declines. Existing methods address this issue at the component level, targeting either the Questioner or the Solver, and overlook that self-evolution is a tightly coupled system. We propose a holistic framework based on learnable information gain, which measures how much novel, parameterizable information a round provides relative to the previous round. Theoretically, this gain equals the Kullback-Leibler divergence between the two rounds' data distributions plus their entropy change. Practically, it is estimated by fitting a small language model to the previous round and scoring new data via negative log-likelihood. Based on this diagnostic, we propose ATRI (Adaptive Training Regulation via Information-gain), which reweights samples within a round and halts training across rounds when information gain remains low. Experiments on popular datasets demonstrate the superiority of our proposal.

cs.CL↗

Knowledge-Intensive Video Generation

Text-to-video generation has advanced rapidly in visual quality, but remains under-evaluated for factuality and practical usefulness. We introduce knowledge-intensive video generation (KIVI), where models generate videos from short information-seeking prompts that ask for explanations, procedures, or demonstrations. To evaluate this setting, we construct KIVI-Bench, a benchmark of 1,080 prompts, and propose automatic metrics for factuality and helpfulness. Human evaluation shows that our metrics significantly better align with human annotations than existing alternatives. Experiments on seven state-of-the-art video generation models show that current systems still lag behind human performance, especially on visual properties, procedural operations, and clear information presentation. These results highlight KIVI as a challenging direction for factual and instructionally useful video generation.

cs.CV↗

In-Context Learning for Robots: Methods and Applications

General-purpose robots must infer what a new task requires and translate that understanding into appropriate physical action. In-context learning (ICL) for robots supports this process by using demonstrations and interaction to direct existing competence with neural parameters held fixed during deployment. We organize this literature review around the interfaces connecting contextual evidence to execution, distinguishing four families: context-conditioned policies, geometric demonstration transfer, world-model-based control, and skill- and agent-based execution. Comparing these interfaces clarifies their transfer assumptions and the roles of training, correspondence, and memory in making context useful. Across manipulation and navigation, we examine how these mechanisms preserve taught requirements as objects, environments, and execution conditions change. This analysis links method design to evaluation practices that distinguish responsiveness to teaching, physical transfer, and benefits from retained experience. The resulting agenda connects compositional task acquisition and faithful transfer with physical recursive self-improvement, in which experience improves the ability to learn subsequent tasks.

cs.RO↗

Robo-Harness K1: Harnessing Robot-Use Agents via Perception Augmentation

Foundation vision-language models (VLMs) understand objects, instructions, and spatial relations, yet translating this capability into robotic manipulation remains difficult. Vision-language-action (VLA) models require extensive demonstrations and may compromise pretrained understanding, while direct RGB-only VLM control is costly and strongly dependent on model capability. We introduce Robo-Harness K1, a robot-use agent (RUA) framework that exposes perception as tools. The agent queries calibrated depth, persistent visual anchors, spatial measurements, and grasp hypotheses, then selects generic motions from the returned evidence. This interface makes 3D geometry accessible without changing the VLM architecture or training a depth encoder. On matched LIBERO-PRO tasks, Gemini 3.7 Flash with K1 reaches 77.8% accuracy, surpassing GPT-6 Astra's 61.1% with an RGB-only harness; K1 further improves Astra to 88.9%. Without target fine-tuning, Gemini with K1 transfers to three RoboSuite arms and dual-arm RoboTwin tasks. On RoboTwin, it achieves 32.0% on Easy and 28.0% on Hard, showing resilience to visual and environmental perturbations. K1 also produces tool-call traces aligned with next-token training. A Qwen3.5-9B student trained on only 107 teacher episodes reaches 44.2% accuracy on new initial states versus 30.2% for OpenVLA, and 13.9% on held-out task conditions versus 0.0% for OpenVLA. These results suggest that perception-augmented RUAs offer a promising route to sample-efficient, generalizable robotic policies that leverage VLM capabilities through an accessible tool interface.

cs.RO↗

RACaP: Agentic Reasoning, Acting, and Coding as Policies for Evolvable Robot Learning

General-purpose robot agents must learn from experience, transfer to new tasks, and act efficiently. Code as Policies (CaP) methods generate and repair programs at runtime, incurring latency and entangling reusable mechanisms with task-specific decisions. We introduce RACaP, an agentic framework that moves coding to evolution and uses a Reasoning-and-Acting (ReAct) loop to call frozen, typed Policy APIs at deployment. A two-phase strategy combines capability curriculum learning with autonomous self-evolution to improve the APIs, the ReAct harness, and experience memory. The APIs encode reusable physical mechanisms while exposing arguments for runtime adaptation. ReAct combines task-specific working memory, long-term experience memory, and visual feedback to select actions, verify outcomes, and recover from failures without modifying source code. RACaP achieves 54.4% success on LIBERO-90, 45.0% on zero-shot LIBERO-PRO, and 46.0% on LIBERO-Long, compared with at most 4.0% for CaP baselines on long-horizon tasks. On LIBERO-PRO, it achieves 2.5 times the success rate of CaP baselines and a 1.9-fold speedup in median policy time. For efficient on-robot deployment, rejection-sampled fine-tuning distills GPT-5.6 ReAct decisions into Qwen3-VL-8B-Instruct, yielding a 13.2-fold per-decision inference speedup and reducing repeated physical calls from 16 to 4. These results show that separating reusable code from runtime decisions supports continued evolution, effective transfer, and efficient long-horizon control.

cs.RO↗

World Action Agent: Harnessing VLMs for Robot Manipulation via World Action Rehearsal

General-purpose vision-language models (VLMs) bring broad knowledge and spatial reasoning to robot manipulation, yet existing systems either use them indirectly, to predict constraints or write programs, or give them a view of the scene rather than a world in which to act. We present World Action Agent (WAA), a multi-agent harness through which VLMs pilot robots with basic tools, making every decision within a visual action workspace. The workspace has three properties. Contact views, selected automatically from the scene geometry, present the scene around the current interaction. Action rehearsal turns each action into an editable proposal that the agent, alone or through an Imagination Agent, previews and revises against planning feedback before execution. In-view correction closes the loop between observation, rehearsal, and low-level execution, letting the agent remove residual offsets in the view where it observes them. Through the same workspace, WAA acquires embodied procedural knowledge in two ways: it evolves multimodal skills from expert videos and human teaching under evidence-based review and consults them through a Skill Agent, and its interaction traces train smaller VLMs to pilot the same harness. On LIBERO-Pro, WAA with skills evolved only from LIBERO-90 reaches a state-of-the-art 75.6% average success, outperforming end-to-end VLAs, code-as-policy agents, and a visual-harness baseline with the same backbone; the same skills remain effective on robosuite without further learning. Fine-tuning Qwen3.5-9B on harness traces raises its out-of-domain success from 1.7% to 43.3%.

cs.RO↗

VeriTS: Verifiable Model-Enhanced Time-Series Queries on Blockchain Systems

Every blockchain transaction carries a timestamp, and the chain imposes a total order. On-chain data therefore forms per-source time-series streams. However, existing systems support only basic lookups on blocks and transactions, and cannot answer time-series queries such as time-range retrieval and windowed aggregation. Offloading queries off-chain restores expressiveness, but the off-chain query layer is untrusted, so results must be verifiable. To this end, we propose VeriTS, the first verifiable time-series query framework for blockchain systems. It supports efficient range and aggregation queries. VeriTS maintains an off-chain query layer. In this layer, each stream is kept under one tree whose nodes carry authenticated aggregates, so the query index is itself the authenticated data structure. A light client thus verifies a windowed aggregate from a logarithmic number of authenticated nodes rather than from every record. VeriTS further answers error-tolerant queries from compact model representations of a stream, and extends the completeness and soundness guarantees to such approximate answers. As VeriTS never trusts the model behind a representation, a faulty or adversarial model can only widen an answer's certified interval, never falsify it. Experiments offer evidence that on windowed aggregation, VeriTS improves verification efficiency by more than two orders of magnitude over per-record proofs. On range retrieval, proofs shrink by up to 14.5x.

cs.DB↗