BZPEERReading, trust and collaboration for research
Menu

Discover

GraphFollow connections around a paper, topic or saved view.

Workspaces

HomePick up where your research flow left off.InboxHandle requests, replies and notifications from one place.CollectionsCurate reading lists, evidence sets and shareable dossiers.ResearchSee your private provenance graph, trail and imports.CollaborationMove from discovery into focused discussion rooms.PublishDraft, preview and publish full-text research articles.

Network

ExpertsFind specialists worth following or asking for help.CommunitiesJoin research circles organized around shared work.PartnersSee external organizations active around the same topics.

Opportunities

ListingsBrowse roles, calls and collaborations with clearer fit signals.

Account

Log inReturn to your reading and collaboration workspace.Create accountStart saving work, following people and joining teams.
Log inCreate account

Researcher profile

Zhongwei Wan

Zhongwei Wan contributes to research discovery and scholarly infrastructure.

ResearcherAffiliation not importedOpen to collaborate
Computation and LanguageNeural and Evolutionary ComputingComputer Vision

Trust snapshot

Quick read

Trust 19 - UnverifiedVerification L1Unclaimed author
5works
0followers
3topics
4close collaborators

Actions

Decide how to stay connected

Follow researcher0

Identity and collaboration

How to connect with this researcher

Claiming links this public author record to a researcher profile and unlocks direct collaboration workflows.

Log in to claim

Direct collaboration

Open a focused conversation when the fit is right

Claim this author entity first to unlock direct invitations.

Research graph

See the researcher in context

Open full explorer

Inspect adjacent work, topics, institutions and collaborators without jumping out to a separate graph page.

Building this graph slice

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Published work

5 published item(s)

preprint2026arXiv

MMDeepResearch-Bench: A Benchmark for Multimodal Deep Research Agents

Deep Research Agents (DRAs) generate citation-rich reports via multi-step search and synthesis, yet existing benchmarks mainly target text-only settings or short-form multimodal QA, missing end-to-end multimodal evidence use. We introduce MMDeepResearch-Bench (MMDR-Bench), a benchmark of 140 expert-crafted tasks across 21 domains, where each task provides an image-text bundle to evaluate multimodal understanding and citation-grounded report generation. Compared to prior setups, MMDR-Bench emphasizes report-style synthesis with explicit evidence use, where models must connect visual artifacts to sourced claims and maintain consistency across narrative, citations, and visual references. We further propose a unified, interpretable evaluation pipeline: Formula-LLM Adaptive Evaluation (FLAE) for report quality, Trustworthy Retrieval-Aligned Citation Evaluation (TRACE) for citation-grounded evidence alignment, and Multimodal Support-Aligned Integrity Check (MOSAIC) for text-visual integrity, each producing fine-grained signals that support error diagnosis beyond a single overall score. Experiments across 25 state-of-the-art models reveal systematic trade-offs between generation quality, citation discipline, and multimodal grounding, highlighting that strong prose alone does not guarantee faithful evidence use and that multimodal integrity remains a key bottleneck for deep research agents.

0 citations0 reviewsNo code linkedOpen access
preprint2026arXiv

MMFormalizer: Multimodal Autoformalization in the Wild

Autoformalization, which translates natural language mathematics into formal statements to enable machine reasoning, faces fundamental challenges in the wild due to the multimodal nature of the physical world, where physics requires inferring hidden constraints (e.g., mass or energy) from visual elements. To address this, we propose MMFormalizer, which extends autoformalization beyond text by integrating adaptive grounding with entities from real-world mathematical and physical domains. MMFormalizer recursively constructs formal propositions from perceptually grounded primitives through recursive grounding and axiom composition, with adaptive recursive termination ensuring that every abstraction is supported by visual evidence and anchored in dimensional or axiomatic grounding. We evaluate MMFormalizer on a new benchmark, PhyX-AF, comprising 115 curated samples from MathVerse, PhyX, Synthetic Geometry, and Analytic Geometry, covering diverse multimodal autoformalization tasks. Results show that frontier models such as GPT-5 and Gemini-3-Pro achieve the highest compile and semantic accuracy, with GPT-5 excelling in physical reasoning, while geometry remains the most challenging domain. Overall, MMFormalizer provides a scalable framework for unified multimodal autoformalization, bridging perception and formal reasoning. To the best of our knowledge, this is the first multimodal autoformalization method capable of handling classical mechanics (derived from the Hamiltonian), as well as relativity, quantum mechanics, and thermodynamics. More details are available on our project page: MMFormalizer.github.io

0 citations0 reviewsNo code linkedOpen access
preprint2026arXiv

PRISM: Pareto-Efficient Retrieval over Intent-Aware Structured Memory for Long-Horizon Agents

Long-horizon language agents accumulate conversation history far faster than any fixed context window can hold, making memory management critical to both answer accuracy and serving cost. Existing approaches either expand the context window without addressing what is retrieved, perform heavy ingestion-time fact extraction at substantial token cost, or rely on heuristic graph traversal that leaves both accuracy and efficiency on the table. We present PRISM, a training-free retrieval-side framework that treats long-horizon memory as a joint retrieval-and-compression problem over a graph-structured memory. PRISM combines four orthogonal inference-time components: Hierarchical Bundle Search over typed relation paths, Query-Sensitive Edge Costing that aligns traversal with detected query intent, Evidence Compression that compresses the candidate bundle into a compact answer-side context, and Adaptive Intent Routing that routes most queries through zero-LLM tiers. By formulating retrieval as min-cost selection over typed path templates and pairing it with an LLM-side compression step, PRISM surfaces the right evidence under a strict context budget without any fine-tuning or modification to the upstream ingestion pipeline. Experiments on the LoCoMo benchmark show that PRISM delivers substantially higher LLM-judge accuracy than every same-protocol baseline at an order-of-magnitude smaller context budget, occupying a previously empty corner of the accuracy-context-cost frontier and demonstrating a superior balance between answer quality and retrieval efficiency.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

Large Scale Many-Objective Optimization Driven by Distributional Adversarial Networks

Estimation of distribution algorithms (EDA) as one of the EAs is a stochastic optimization problem which establishes a probability model to describe the distribution of solutions and randomly samples the probability model to create offspring and optimize model and population. Reference Vector Guided Evolutionary (RVEA) based on the EDA framework, having a better performance to solve MaOPs. Besides, using the generative adversarial networks to generate offspring solutions is also a state-of-art thought in EAs instead of crossover and mutation. In this paper, we will propose a novel algorithm based on RVEA[1] framework and using Distributional Adversarial Networks (DAN) [2]to generate new offspring. DAN uses a new distributional framework for adversarial training of neural networks and operates on genuine samples rather than a single point because the framework also leads to more stable training and extraordinarily better mode coverage compared to single-point-sample methods. Thereby, DAN can quickly generate offspring with high convergence regarding the same distribution of data. In addition, we also use Large-Scale Multi-Objective Optimization Based on A Competitive Swarm Optimizer (LMOCSO)[3] to adopts a new two-stage strategy to update the position in order to significantly increase the search efficiency to find optimal solutions in huge decision space. The propose new algorithm will be tested on 9 benchmark problems in Large scale multi-objective problems (LSMOP). To measure the performance, we will compare our proposal algorithm with some state-of-art EAs e.g., RM-MEDA[4], MO-CMA[10] and NSGA-II.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

Many-Objective Estimation of Distribution Optimization Algorithm Based on WGAN-GP

Estimation of distribution algorithms (EDA) are stochastic optimization algorithms. EDA establishes a probability model to describe the distribution of solution from the perspective of population macroscopically by statistical learning method, and then randomly samples the probability model to generate a new population. EDA can better solve multi-objective optimal problems (MOPs). However, the performance of EDA decreases in solving many-objective optimal problems (MaOPs), which contains more than three objectives. Reference Vector Guided Evolutionary Algorithm (RVEA), based on the EDA framework, can better solve MaOPs. In our paper, we use the framework of RVEA. However, we generate the new population by Wasserstein Generative Adversarial Networks-Gradient Penalty (WGAN-GP) instead of using crossover and mutation. WGAN-GP have advantages of fast convergence, good stability and high sample quality. WGAN-GP learn the mapping relationship from standard normal distribution to given data set distribution based on a given data set subject to the same distribution. It can quickly generate populations with high diversity and good convergence. To measure the performance, RM-MEDA, MOPSO and NSGA-II are selected to perform comparison experiments over DTLZ and LSMOP test suites with 3-, 5-, 8-, 10- and 15-objective.

0 citations0 reviewsNo code linkedOpen access