arXiv subjects

Yuen-Hei Yeung

Publications and source records attributed to Yuen-Hei Yeung.

6 recordsLinked to original sources

arXiv2026

Learning Scene-Aware Humanoid Locomotion through 3D Clutter from Immersive Human Demonstrations

While learning from human motions has enabled highly dynamic humanoid skills such as dancing and martial arts in obstacle-free space, traversal through densely cluttered environments remains underexplored. These spaces are three-dimensional and geometrically constrained, requiring scene-aware locomotion that tightly couples whole-body motion with scene geometry for obstacle avoidance. To address these challenges, we present Moving Through Clutter (MTC), a learning-from-demonstration framework for scene-aware humanoid locomotion. To bypass costly physical scene construction, MTC uses procedurally generated Virtual Reality environments for immersive data collection. To transform these human motions into training-ready humanoid motions, we propose a scene-aware motion retargeting algorithm that converts human demonstrations into humanoid trajectories while strictly enforcing robot-scene clearance to guarantee collision-free traversal. These reference trajectories are then used to train a scene-aware locomotion policy that deploys on a Unitree G1 humanoid. Evaluated on our proposed MTC-Challenge for multi-obstacle traversal, the policy demonstrates a 70.2% collision-free rate across diverse scenarios, successfully traversing complex environments through diverse whole-body skills, including crawling through low-clearance passages and squeezing through narrow gaps.

cs.RO↗

arXiv2026

Demonstration Synthesis from a Single Scan via Gaussian Splatting for Visuomotor Policy Learning

Training a visuomotor policy calls for abundant demonstrations that closely match the target environment, yet collecting them anew remains expensive. Existing demonstration synthesis methods reduce this cost but remain constrained by high manual effort, limited visual fidelity, or heavy reliance on physics simulators. This paper introduces GaussianFactory, a high-fidelity data engine that mass-produces demonstrations with a single video scan as its only human input and no physics engine in the generation loop. Specifically, GaussianFactory reconstructs the scene as an editable 3D Gaussian Splatting (3DGS) replica and samples from the object-combination tasks the scene affords. For each task, it plans grasps and trajectories purely kinematically on the geometric reconstruction, rendering photorealistic demonstrations that visually match the target environment. Physical dynamics enter the pipeline only where contact force interactions dictate the outcome---during grasp formation, via a learned contact model pretrained once on an interaction dataset. To evaluate the downstream utility of the synthesized demonstrations, we implement an end-to-end scan-to-deployment workflow in two setups: a simulated scene that stands in for the real world to enable reproducibility, and a real-world workspace with a physical UR10e robot. In each setup, a standard diffusion policy trained solely on the synthesized demonstrations achieves 95.1% and 84.2% success rates, respectively.

cs.RO↗

arXiv2026

Resist, Update, Reject: Preference Optimization Installs a Prior-Dependent Reliability Switch

An aligned model asked to hold its answer against a manipulative source must still update on a reliable one and reject an unreliable one: resistance, reliable-update, and unreliable-source rejection are one three-way contract, not three independent behaviors. We show the objective most anti-sycophancy work optimizes is non-identifying with respect to source reliability: because no preference label depends on whether a source is actually reliable, any scalar mixture of the arms traces a single deference dial, and no point separates two same-template testimonies differing only in stated reliability. This fixation$\leftrightarrow$gullibility frontier is a property of the objective, not any model. We make reliability identifiable through data: a threshold benchmark where a source asserts the opposite answer while stating its reliability $r$, and the correct action is to flip iff $r$ exceeds the model's prior strength $p$. Preference optimization over balanced coverage installs a prior-dependent reliability switch: across three seeds on Qwen2.5-7B-Instruct the threshold $r^\star$ rises monotonically with the prior, decision accuracy reaches $0.84$ with a monotone flip curve (Spearman $0.56$), and the policy generalizes to unseen reliability values and a held-out notation, following stated reliability over role prestige. Three controls localize the cause: an unmatched variant installs the switch equally ($0.80$), a second preference optimizer (IPO) installs it just as well ($0.86$), whereas supervised imitation does not ($0.50$), so the cause is preference optimization over reliability-labeled coverage, not pairing, loss, or imitation. A confirmatory battery replicates the switch on a fresh test draw, bounds it honestly (it keys on reliability stated in the testimony, not a separately audited record), and transfers it to Llama-3.1-8B. The frontier is empirical, not a theorem.

cs.LG↗

arXiv2026

Resist and Update: Counterfactual Report Coordinates for Incentive-Compatible LLMs

Aligned language models routinely misreport under non-evidential pressure: they cave to a confident user, yet fail to revise when genuine evidence arrives. We cast this as a failure of internal incentive-compatibility and study the two demands, resist (ignore forbidden pressure) and update (follow licensed evidence), on a Bayesian-witness benchmark with known posteriors, where the same user disagreement is evidence or pressure purely by stated source reliability, removing the evidence/pressure confound by construction. Using interchange interventions rather than probes, we causally localize low-rank report coordinates for answer, confidence, and caveat, establishing causal sufficiency at a late intervention site rather than uniqueness or necessity, with a causal cross-talk matrix showing strong own-coordinate control and only small cross-effects (partial functional disentanglement). We then introduce a training-free counterfactual report-coordinate (CRC) clamp that references the model's own report under an incentive-neutralized counterfactual of the prompt. The two-pass full-window clamp attains resist and update of $1.00$ jointly (Wilson 95% CI $[0.99,1.00]$; the rank-16 projection alone reaches $0.88/0.90$), which we read as a causal certificate and upper bound under a constructible reference, not a claim of a deployed solution. Tested global decoding and fixed-direction steering trade one objective against the other, and resist-only training collapses updating to $0.01$. The deployable single-pass compilation is lossy ($0.73/0.97$). The mechanism and the clamp reproduce across three model families and transfer to a natural sycophancy benchmark with significant paired improvements. Our contribution is the interface and certification method: activation-level counterfactual incentive-invariance as a structural primitive for internal incentive-compatibility.

cs.AI↗

arXiv2026

Unlearning as Distribution Restoration: A Controlled Counterfactual Study, a Validated Selective Screen, and the Limits of Oracle-Free Certification

Machine unlearning is commonly evaluated by matching a retrained oracle on trained probes. In a controlled nonce-fact testbed with a matched retraining reference, we find this criterion can favor methods that retain held-out knowledge: candidates it rates adequate score held-out forget facts $-2.82$ nats below the never-learned level (cluster CI $[-3.16,-2.48]$). We recast unlearning as restoration to the matched reference and audit oracle-free screens and certificate-style criteria across 45 model-seed cells spanning five open architecture families. The reference itself falsifies an absolute retain/round-trip certificate: the injected model, which retains the retain set by construction, fails the fixed retain threshold in 41/45 cells and its own round trip in 31/45, and the reference fully certifies in only 1/45. A base-anchored held-out screen remains strong as a selective necessary test: on a sealed challenge suite it rejects the injected model in 45/45 cells, accepts the reference in 44/45, and partially detects entity-routing suppression (35/45); it is a necessary test with measured sensitivity, not a sufficiency certificate. A damage-relative recalibration anchored to the reference's own operating point certifies a small subset in 15/45 cells; where it does not abstain, its picks lie within retraining noise (0.80 nats) on the axes it optimizes, while the common trained-probe criterion sits 5.17 nats away (a supporting comparison, not a head-to-head benchmark). A fixed-magnitude logit-suppression attack defeats the full forward battery in 12/45 cells, so forward-only certification is not sound; our method is an empirical selective test for methods-as-produced. An identifiability theorem delimits which facts admit an oracle-free forget threshold at all, with TOFU as the predicted boundary case.

cs.LG↗

arXiv2024

Leveraging RGB-D Data with Cross-Modal Context Mining for Glass Surface Detection

Glass surfaces are becoming increasingly ubiquitous as modern buildings tend to use a lot of glass panels. This, however, poses substantial challenges to the operations of autonomous systems such as robots, self-driving cars, and drones, as these glass panels can become transparent obstacles to navigation. Existing works attempt to exploit various cues, including glass boundary context or reflections, as priors. However, they are all based on input RGB images. We observe that the transmission of 3D depth sensor light through glass surfaces often produces blank regions in the depth maps, which can offer additional insights to complement the RGB image features for glass surface detection. In this work, we first propose a large-scale RGB-D glass surface detection dataset, \textit{RGB-D GSD}, for rigorous experiments and future research. It contains 3,009 images, paired with precise annotations, offering a wide range of real-world RGB-D glass surface categories. We then propose a novel glass surface detection framework combining RGB and depth information, with two novel modules: a cross-modal context mining (CCM) module to adaptively learn individual and mutual context features from RGB and depth information, and a depth-missing aware attention (DAA) module to explicitly exploit spatial locations where missing depths occur to help detect the presence of glass surfaces. Experimental results show that our proposed model outperforms state-of-the-art methods.

cs.CV↗