Searcharxiv⌕ Search

arXiv subjects

Junhao Cai

Publications and source records attributed to Junhao Cai.

At least 19 recordsLinked to original sources

Retain-Neutral Surrogates for Min-Max Unlearning

Machine unlearning aims to make a model forget designated training data while retaining performance on the rest. In min-max unlearning, an inner adversary perturbs the model to maximize the forget loss, and the model is updated to minimize the retain loss at the resulting surrogate point. When the forget and retain gradients are positively coupled, the standard perturbation also moves toward higher retain loss, and the retain step begins by recovering that loss. We propose Retain-Orthogonal Surrogate Unlearning (ROSU), which redefines the adversary to admit only perturbations orthogonal to the retain gradient. The one that raises the forget loss fastest has a closed form, and ROSU uses it to set the surrogate point and amplifies it into the forgetting step. Under positive coupling the ROSU surrogate avoids the retain cost that coupling induces in the standard surrogate, and the difference shrinks as the gradients approach orthogonality. ROSU has the smallest accuracy gap to retraining in every vision setting (CIFAR-10, CIFAR-100, Tiny-ImageNet) and is competitive with the baselines on language benchmarks (TOFU, WMDP). The redefinition helps most under high positive coupling. Code is available at https://github.com/unlearn-ml/machine-unlearning.

cs.LG↗

Data-Efficient Unlearning-Based Backdoor Detection in Federated Learning

Detecting backdoor attacks in federated learning (FL) is challenging because malicious clients can manipulate the global model while preserving main-task performance. Static model-update statistics may therefore fail to clearly distinguish malicious from benign clients. Unlearning-based detectors instead probe client models through loss ascent on clean server data, but detection performance degrades with scarce proxy data. We propose Screening via Unlearning Response Analysis (SURA), a data-efficient detector. SURA reuses the same small clean proxy across multiple train-mode loss-ascent steps, strengthening client responses without additional clean data. For federated vision models with batch normalization (BN), repeated probing also changes BN running statistics and can confound endpoint measurements. SURA addresses this by restoring the shared pre-probe BN state only at readout while preserving the parameter changes induced by probing. This separates repeated probing from response measurement and removes BN drift and the interaction between parameter and BN-state changes from the readout. Across six backdoor attacks, SURA uses only 10 clean server images, outperforms methods under the same data budget, and approaches the robustness obtained with much larger proxy sets.

cs.CR↗

Spectral Unlearning: Transformer Structure-Preserving Updates for Language Models

Machine unlearning in large language models (LLMs) aims to remove specific learned knowledge, capabilities, or behaviors while preserving model utility. Existing methods have mostly focused on the design of unlearning objectives, while largely overlooking the intrinsic matrix structure of transformer weights. In particular, standard optimizers operate on independent scalar coordinates, thereby ignoring this matrix structure. We propose spectral unlearning, an update framework that combines standard unlearning objectives with matrix-structured optimization. We show that the spectral update corresponds to the steepest-descent direction for transformers composed of linear layers with normalized inputs and outputs. Our approach is applied on top of existing unlearning methods and evaluated across complementary axes of memorization, privacy, and utility, summarized via their harmonic mean. Spectral unlearning consistently outperforms a broad range of baselines and closes most of the gap to a retain-trained reference. Code is available at https://github.com/unlearn-ml/spectral-unlearning/.

cs.LG↗

From Predictive Geometry to Action: Learning Complementary Predictive Representations for Embodied Manipulation

Effective manipulation requires a policy to capture not only about the current scene geometry, but also about how that geometry will evolve under interaction. Existing approaches either augment policies with geometric information from current observations or introduce future prediction as auxiliary supervision, leaving open how predictive geometry should be represented for action generation. Therefore, we propose a two-stage framework, which learns complementary predictive representations from multi-view video clips. During pretraining, a frozen 3D foundation model provides dense geometric features that are compressed into predictive geometric representations and optimized to preserve scene structure and its temporal evolution. Action supervision is then applied to geometric transitions to extract a complementary motor representation that emphasizes action-relevant transitions. During downstream post-training, predictive geometry is transferred through geometry queries, while the motor representation guides intermediate features along the action-generation pathway. Future observations are used only as privileged supervision during training and are not required at inference time. Across LIBERO, LIBERO-Plus, SimplerEnv, and real-world bimanual and single-arm tasks, our approach consistently improves downstream policies, including an absolute gain of 10.4 percentage points on the WidowX evaluation. We further demonstrate transfer across multiple VLA architectures and WAMs, with improvements in both future-world and action prediction. The code will be available at https://github.com/xuxiaoxxxx/CPR.

cs.RO↗

CoRef-GS: Cooperative Referring Gaussian Splatting for Multi-Agent Scene Understanding

Referring scene understanding for embodied robots requires grounding object- and relation-centric language queries from a designated viewpoint. While a local semantic Gaussian map can support such grounding within one agent's observations, cooperative settings require this ability to remain effective after independently reconstructed maps are aligned and fused. In this setting, the referred target or its contextual landmark may come from another agent's observations, while spatial relations must still be interpreted from the querying robot's viewpoint. We formulate this problem as cooperative referring Gaussian grounding over fused maps, which requires geometric alignability, instance-level semantic comparability, and view-conditioned relation reasoning. Existing language-aware Gaussian methods mainly focus on single-map querying, whereas Gaussian registration methods optimize geometric or photometric alignment without preserving language-grounding-oriented semantic compatibility. We propose CoRef-GS, a cooperative referring Gaussian splatting framework. CoRef-GS constructs local open-vocabulary instance-aware Gaussian maps, then aligns partially overlapping maps with a cross-agent alignment module by geometric and semantic consistency, and grounds queries using a view-conditioned mask relation graph. We further introduce CoQuad-Ref, a dual-quadruped benchmark spanning both real-world and simulated indoor scenes. Experiments show that, on simulated scenes, CoRef-GS reduces the rotation error from 2.58° after coarse initialization to 0.15° after refinement, and improves real-world referring mIoU over ReferSplat from 52.6% to 68.8%. The established benchmark and source code will be publicly released at https://github.com/ruojiruoli17/CoRef-GS.git.

cs.RO↗

Beyond Viewpoint Generalization: What Multi-View Demonstrations Offer and How to Synthesize Them for Robot Manipulation?

Does multi-view demonstration truly improve robot manipulation, or merely enhance cross-view robustness? We present a systematic study quantifying the performance gains, scaling behavior, and underlying mechanisms of multi-view data for robot manipulation. Controlled experiments show that, under both fixed and randomized backgrounds, multi-view demonstrations consistently improve single-view policy success and generalization. Performance varies non-monotonically with view coverage, revealing effective regimes rather than a simple "more is better" trend. Notably, multi-view data breaks the scaling limitation of single-view datasets and continues to raise performance ceilings after saturation. Mechanistic analysis shows that multi-view learning promotes manipulation-relevant visual representations, better aligns the action head with the learned feature distribution, and reduces overfitting. Motivated by the importance of multi-view data and its scarcity in large-scale robotic datasets, as well as the difficulty of collecting additional viewpoints in real world settings, we propose RoboNVS, a geometry-aware self-supervised framework that synthesizes novel-view videos from monocular inputs. The generated data consistently improves downstream policies in both simulation and real-world environments.

cs.RO↗

XPolicyLab: A Unified Standard and Open Ecosystem for Robot Policy Evaluation and Deployment

Robot policy evaluation and deployment remain fragmented by model-specific software dependencies, data representations, and runtime interfaces, so that connecting N policies to M evaluation environments requires O(NM) separate integrations. We present XPolicyLab, a unified standard and open ecosystem that reduces this cost to O(N+M). XPolicyLab specifies common observation, action, and trajectory schemas together with a minimal adapter interface for observation updates, action prediction, batched execution, and episode reset, while a dependency-isolated client/server architecture separates policy inference from environment execution, so that each side retains its native software stack and may run locally or remotely. The ecosystem integrates 42 robot policies and standardizes their installation, debugging, serving, and evaluation workflows. Across these adapters, model-specific code varies by an order of magnitude while the environment-facing loop stays within a few lines of a fixed reference, confirming that the contract confines heterogeneity to the policy side. In a controlled study, conforming to the standard reduces the integration effort of a representative policy from over five hours to two hours, and packaged agent skills reduce it further to thirty minutes. The same adapters serve RoboTwin, RoboDojo simulation, and standardized real-robot evaluation through one interface. XPolicyLab is released as shared infrastructure for reproducible policy comparison and standardized deployment across simulation and physical platforms. Project website: https://xpolicylab.github.io/.

cs.RO↗

SCOPE: Entanglement Frontier Escape for Source-Free Class Unlearning

Source-free class unlearning erases whole classes using only the forget data, judged at the representation level, where features can leak a class the head no longer predicts. Existing feature-space erasers answer with one fixed projection, yet forget and retain classes share a representation, so deleting one disturbs the other where they overlap. We prove this tension is a frontier. Every fixed projection that deletes pays a retain cost of at least the retain-readout energy along the forget-discriminant subspace, and erasing that subspace alone attains the floor. The leading source-free erasers all instantiate the form it binds, so the frontier limits the whole class. Conditioning the erasure on the input escapes it. Spectral Conditional Projective Erasure (SCOPE) does so with a single gate, suppressing the forget subspace chiefly on inputs its frozen head's weight scores read as a forget class. It is closed form, needs no retain data or gradient training, and costs orders of magnitude less than retraining. Across five object, face, and speaker benchmarks spanning two modalities and both convolutional and transformer backbones, the frontier predicts the measured retain cost. SCOPE leads the source-free erasers on every benchmark and forget-set size, and at the hardest setting it tops every unlearner, trained methods included.

cs.LG↗

RoboInter1.5: A Holistic Intermediate Representation Suite for Embodied World Modeling and Robotic Manipulation

Existing robot datasets remain expensive to curate, embodiment-specific, and insufficiently annotated with the fine-grained structure required for generalizable reasoning, execution, or long-horizon environment dynamics simulation. Building on our prior work, RoboInter1.0, we present RoboInter1.5, an extended and holistic suite of intermediate representations for both robotic manipulation and embodied world modeling. RoboInter1.5 provides a unified resource of data, benchmarks, and models centered on dense manipulation-oriented intermediate representations. Specifically, RoboInter-Data contains over 230k manipulation episodes across 571 scenes with dense per-frame annotations covering more than ten types of intermediate representations, including subtasks, primitive skills, object and gripper grounding, segmentation, affordance, grasp poses, contact points, motion traces, etc. Built upon these annotations, RoboInter-VQA introduces spatial and temporal embodied VQA tasks to benchmark and improve the intermediate-representation reasoning capabilities of our RoboInter-VLM. RoboInter-VLA further studies how such representations benefit action execution through implicit, explicit, and modular plan-then-execute paradigms. To better model the physical world, we further introduce RoboInter-World, which leverages intermediate representations as structured conditioning signals for controllable prediction of future world states. Extensive evaluations demonstrate that RoboInter1.5 provides a unified spatiotemporal scaffolding for intermediate representations. Rather than treating intermediate representations merely as interpretable signals, RoboInter1.5 conceptualizes them as a bidirectional interface that both regularizes low-level action spaces and constrains the latent rollouts of open-world physical simulators.

cs.RO↗

InternVLA-A1.5: Unifying Understanding, Latent Foresight, and Action for Compositional Generalization

Unified models for robot manipulation aim to equip one policy with both the semantic priors of pretrained VLMs and the physical dynamics learned through future prediction. In practice, existing designs tend to erode the semantics of the pretrained backbone, suffer interference among heterogeneous objectives, and learn future prediction from scratch in pixel space, leaving the dynamics priors of pretrained video generators unexploited. We present InternVLA-A1.5, which builds the policy on a native VLM backbone that keeps training on VQA and subtask prediction, and attaches a lightweight unified expert for continuous action generation. Future prediction is recast as a latent-querying problem, where a small set of learnable foresight tokens condenses the task-relevant future into a compact latent code under the supervision of a frozen pretrained video generation model, so the policy inherits world-model dynamics priors without ever learning pixel-level generation. The video branch is discarded at inference, keeping real-time control. Pretrained on 1.2M robot episodes and 3M multimodal samples, InternVLA-A1.5 achieves the best overall results on all six simulation benchmarks. In the real world, the preserved semantics deliver the strongest compositional generalization on held-out instruction bindings, and the two designs together sustain long-horizon execution.

cs.RO↗

CMDS-AD: Cross-Modal Dual-Stream Decoupling for Few-Shot Anomaly Detection

Few-shot anomaly detection remains challenging due to limited training data. Multi-modal anomaly detection (MAD) offers a viable solution, leveraging 3D geometric cues to enrich 2D RGB representations and compensate for this scarcity. However, existing MAD methods apply spatially uniform feature processing, conflating stable macroscopic structures with high-frequency localized defect signals, exacerbating cross-modal misalignment and inflating false-positive rates. To overcome this, we present CMDS-AD, a Cross-Modal Dual-Stream Anomaly Detection framework. A LoRA-guided diffusion model generates diverse RGB samples to mitigate extreme data scarcity. For 3D normal augmentation, we employ a pre-trained diffusion model as a normal estimator. Crucially, this estimator inherently acts as a non-linear low-pass filter, directly extracting low-frequency normal representations from RGB inputs. This establishes an auxiliary estimated stream of purely low-frequency information, anchoring robust structural templates and assisting the uncompressed real stream, containing coupled high- and low-frequency components, to precisely isolate micro-defects. A Coordinate-Aware Hierarchical Feature Mapper adaptively aligns cross-modal semantics, while a multiplicative scoring mechanism filters modality-specific noise. Under the extreme 1-shot setting, CMDS-AD achieves absolute performance gains of 5.7% (I-AUROC) and 2.0% (AUPRO) on MVTec 3D-AD, alongside 7.7% and 5.6% improvements on EyeCandies, establishing a new state-of-the-art. Code is available at https://github.com/Junhaocai27/CMDS-AD

cs.CV↗

ForgeDreamer: Industrial Text-to-3D Generation with Multi-Expert LoRA and Cross-View Hypergraph

Current text-to-3D generation methods excel in natural scenes but struggle with industrial applications due to two critical limitations: domain adaptation challenges where conventional LoRA fusion causes knowledge interference across categories, and geometric reasoning deficiencies where pairwise consistency constraints fail to capture higher-order structural dependencies essential for precision manufacturing. We propose a novel framework named ForgeDreamer addressing both challenges through two key innovations. First, we introduce a Multi-Expert LoRA Ensemble mechanism that consolidates multiple category-specific LoRA models into a unified representation, achieving superior cross-category generalization while eliminating knowledge interference. Second, building on enhanced semantic understanding, we develop a Cross-View Hypergraph Geometric Enhancement approach that captures structural dependencies spanning multiple viewpoints simultaneously. These components work synergistically improved semantic understanding, enables more effective geometric reasoning, while hypergraph modeling ensures manufacturing-level consistency. Extensive experiments on a custom industrial dataset demonstrate superior semantic generalization and enhanced geometric fidelity compared to state-of-the-art approaches. Code is available at https://github.com/Junhaocai27/ForgeDreamer

cs.CV↗

InternVLA-A1: Unifying Understanding, Generation and Action for Robotic Manipulation

Prevalent Vision-Language-Action (VLA) models are typically built upon Multimodal Large Language Models (MLLMs) and demonstrate exceptional proficiency in semantic understanding, but they inherently lack the capability to deduce physical world dynamics. Consequently, recent approaches have shifted toward World Models, typically formulated via video prediction; however, these methods often suffer from a lack of semantic grounding and exhibit brittleness in the presence of video prediction errors. To synergize semantic understanding with dynamic predictive capabilities, we present InternVLA-A1. This model employs a unified Mixture-of-Transformers architecture, coordinating three experts for scene understanding, visual foresight generation, and action execution. These components interact seamlessly through a unified masked self attention mechanism. Building upon InternVL3 and Qwen3-VL, we instantiate InternVLA-A1 at 2B and 3B parameter scales. We pre-train these models on heterogeneous data sources over real-world robot data, synthetic simulation data, and human videos, covering over 692M frames. This hybrid training strategy effectively harnesses the diversity of synthetic simulation data while minimizing the sim-to-real gap. We evaluated InternVLA-A1 on 12 real-world robotic tasks and a simulation benchmark. The results show that InternVLA-A1 consistently outperforms prior leading models: compared with pi0.5, it achieves +4.4\% on static manipulation tasks and +2.6\% on the RoboTwin 2.0 simulation benchmark, and delivers a +26.7\% boost on dynamic manipulation tasks.

cs.RO↗

HCF: Hierarchical Cascade Framework for Distributed Multi-Stage Image Compression

Distributed multi-stage image compression -- where visual content traverses multiple processing nodes under varying quality requirements -- poses challenges. Progressive methods enable bitstream truncation but underutilize available compute resources; successive compression repeats costly pixel-domain operations and suffers cumulative quality loss and inefficiency; fixed-parameter models lack post-encoding flexibility. In this work, we developed the Hierarchical Cascade Framework (HCF) that achieves high rate-distortion performance and better computational efficiency through direct latent-space transformations across network nodes in distributed multi-stage image compression systems. Under HCF, we introduced policy-driven quantization control to optimize rate-distortion trade-offs, and established the edge quantization principle through differential entropy analysis. The configuration based on this principle demonstrates up to 0.6dB PSNR gains over other configurations. When comprehensively evaluated on the Kodak, CLIC, and CLIC2020-mobile datasets, HCF outperforms successive-compression methods by up to 5.56% BD-Rate in PSNR on CLIC, while saving up to 97.8% FLOPs, 96.5% GPU memory, and 90.0% execution time. It also outperforms state-of-the-art progressive compression methods by up to 12.64% BD-Rate on Kodak and enables retraining-free cross-quality adaptation with 7.13-10.87% BD-Rate reductions on CLIC2020-mobile.

cs.CV↗

InternData-A1: Pioneering High-Fidelity Synthetic Data for Pre-training Generalist Policy

Recent works explore how real and synthetic data contribute to Vision-Language-Action (VLA) models' generalization. While current VLA models have shown the strong effectiveness of large-scale real-robot pre-training, synthetic data has not previously demonstrated comparable capability at scale. This paper provides the first evidence that synthetic data alone can match the performance of the strongest $π$-dataset in pre-training a VLA model, revealing the substantial value of large-scale simulation. The resulting model also exhibits surprisingly zero-shot sim-to-real transfer on several challenging tasks. Our synthetic dataset, InternData-A1, contains over 630k trajectories and 7,433 hours across 4 embodiments, 18 skills, 70 tasks, and 227 scenes, covering rigid, articulated, deformable, and fluid-object manipulation. It is generated through a highly autonomous, fully decoupled, and compositional simulation pipeline that enables long-horizon skill composition, flexible task assembly, and heterogeneous embodiments with minimal manual tuning. Using the same architecture as $π_0$, we pre-train a model entirely on InternData-A1 and find that it matches the official $π_0$ across 49 simulation tasks, 5 real-world tasks, and 4 long-horizon dexterous tasks. We release the dataset and will open-source the generation pipeline to broaden access to large-scale robotic data and to lower the barrier to scalable data creation for embodied AI research.

cs.RO↗

Recent advances in stimulus-assisted nanoprecipitation for nanoparticle synthesis

Nanoprecipitation, the rapid solvent-displacement route to nanoscale phase separation, has matured from a simple batch operation into a versatile platform for nanomaterial synthesis. This review synthesizes recent progress in stimulus-assisted nanoprecipitation, wherein externally applied triggers (ultrasonic, electrical, supergravity, thermal, chemical, and photonic/other stimuli) are integrated with contemporary mixing technologies (batch, flash, microfluidic, membrane and high-shear reactors) to decouple and selectively control over nucleation, growth kinetics, and assembly processes. These methods allow for the precise tuning of the size, morphology, stability and functionality of nanoparticles (NPs), thereby broadening their applications in drug delivery, catalysis and materials science. We distill mechanistic principles by which each stimulus alters local supersaturation, chain mobility, interfacial instabilities, or droplet/film microreactor dynamics, and compare advantages and limitations by surveying research works from recent years. We also explore the potential development trends of multiscale coupling models, design rules for stimulus-compatible continuous reactors, and adoption of data-driven optimization frameworks to expand the capabilities of nanoprecipitation for advanced nanomaterial design.

physics.chem-ph↗

Tuning laser-induced optical breakdown and cavitation through the ionic environment in aqueous media

Laser-induced cavitation in liquids originates from optical breakdown processes that depend sensitively on both laser-plasma dynamics and the chemical microenvironment of the solvent. Herein, we experimentally decouple the effects of ionic strength and ion specificity on cavitation inception in aqueous electrolytes spanning neutral, acidic, and alkaline regimes. Using focused nanosecond laser pulses, we show that increasing ionic strength universally lowers the cavitation threshold by enhancing charge screening and seed-electron availability. However, under constant ionic strength, strongly asymmetric behavior emerges: acidic (hydrogen chloride, HCl) solutions inhibit cavitation, whereas alkaline (sodium hydroxide, NaOH) solutions enhance it. This asymmetry arises from hydrated-electron kinetics that depends on the ion specificity. In acidic solutions, hydronium ions act as diffusion-limited scavengers of hydrated electrons, quenching their lifetime and inhibiting avalanche ionization. In contrast, hydroxide ions reduce hydronium availability and extend electron survival, promoting more efficient plasma formation. Despite large discrepancies in breakdown thresholds, electrical conductivity remains nearly constant, demonstrating that microscopic electron chemistry, rather than bulk charge transport, governs cavitation onset. These results establish a mechanism connection between electrolyte chemistry and optical breakdown, showing that ion-specific reaction dynamics fundamentally control laser-induced cavitation in aqueous environments.

physics.chem-ph↗

Ionic strength-driven cavitation nucleation: from energy deposition-based to tension-based cavitation

In this work, we present a unified experimental and simulation investigation of cavitation in aqueous electrolyte solutions, combining nanosecond laser-induced optical breakdown and all-atom molecular dynamics (MD) simulations under tensile stress. Across both cavitation scenarios, we find that cavitation inception and intensity (bubble nucleation count, cavitation-zone length, vapor-volume fraction) are governed by ionic strength alone, with negligible dependence on the ion species. In laser experiments, increasing ionic strength lowers the breakdown threshold and amplifies bubble generation by supplying extra seed electrons for inverse Bremsstrahlung-driven avalanche ionization. We elucidate the mechanism of action of the ionic strength through the MD simulations, which essentially quantifies the net charge density in the bulk, and thus its combined influence on the generation of seed electrons and the perturbation of the hydration network. These findings identify ionic strength serving as a unifying parameter controlling cavitation in electrolyte solutions: whether driven by rapid energy deposition or by tensile stress imposed.

physics.flu-dyn↗