arXiv subjects

Ruikun Luo

Publications and source records attributed to Ruikun Luo.

7 recordsLinked to original sources

arXiv2026

Answer-Only Disclosure: Private-Context LLM Assistance over Web APIs with a Leakage Ledger

Web assistants increasingly answer questions about a user's private documents by pairing a small on-device model with a cloud LLM behind a Web API. In collaborative decoding, the document stays on the client, and the cloud receives only a public view that the user approves, such as the question alone, but the two models exchange messages while the answer is generated. We show that the message transcript can disclose private content that neither the public view nor the answer contains. In a planted-secret audit, speculative decoding with private drafts did not include the secret code in its answers, yet sent the code to the cloud in rejected drafts. We present ADLER, a decoding protocol for streaming LLM APIs in which every message to the cloud depends only on the public view and the answer the user sees, regardless of how the client combines the two models or how the cloud behaves. A client-side leakage ledger then bounds how much the answer itself reveals relative to a public reference process built from the cloud's replies. In the audit, no ADLER configuration exposed the secret beyond the answer. On GSM8K math problems with numerical values masked in the cloud's input, ADLER with an unlimited leakage budget raises the exact-match accuracy of a 0.5B client from 0.41 to 0.83 and of a 1.5B client from 0.67 to 0.85 under lenient answer parsing, close to the 0.92 achieved by uploading the full problem. It also exceeds a zero-leakage program baseline (0.69) by 0.14 and 0.16, respectively. In this baseline, the cloud writes a program that the client runs offline using the private numbers. The ledger records the disclosure charge associated with this gain; budgets of 16 nats or less leave accuracy near the cloud-only baseline on GSM8K.

cs.CR↗

arXiv2026

Direct Hidden-State Alignment: Mapping and Controlling Preference Expression in LLMs

In many settings, post-training need not create the target behavior from scratch: the base model can already produce it, but not reliably. This shifts part of preference alignment from capability acquisition to behavioral expression. We ask how a specified preference is represented in native model computation, what prevents target-supporting computation from reliably dominating generation, and whether this structure can directly guide control. We introduce Residual Competition Maps (RCMs), which map a behavioral preference onto signed causal effects of native residual computation. Across preference domains, RCMs reveal coexisting target-supporting and target-competing effects, input-dependent component roles, and cases where a single native-component intervention reverses the preference outcome. DPO substantially reorganizes these effects and can weaken opposition without guaranteeing its removal. We then propose Direct Hidden-State Alignment (DHSA), which treats inference-time hidden states rather than base-model weights as the direct adaptation space. RCM-guided Causal Activation State Transition (CAST) implements DHSA through local state interventions at a small number of preference-relevant interfaces while freezing the base model. With only 256-16,384 controller parameters, CAST reaches DPO-competitive operating points across three preference domains, can complement DPO-trained models, and can be enabled or removed at inference time.

cs.LG↗

arXiv2026

WCCS: Efficient Wedge Conductance Community Search over Large Temporal Bipartite Graphs (Full Paper)

Bipartite graphs are ubiquitous for modeling complex interactions between two distinct entity types across numerous practical applications such as e-commerce, academic networks, and social systems. Despite significant progress in community search over bipartite graphs, most prior work is limited to static settings and ignores the rich temporal dynamics present in real-world networks. Moreover, existing methods typically adopt edge-centric measures and strict consecutivity constraints, failing to capture higher-order interactions and frequent yet non-consecutive activities. More importantly, they often neglect the crucial community-quality requirements of both internal cohesiveness and external sparsity, failing to identify critical nodes or including many irrelevant nodes. To address these dilemmas, we propose the novel problem of \emph{Wedge Conductance Community Search (WCCS)}, which aims to identify a query-dependent community that is not only structurally and temporally cohesive but also well-separated from the rest of the network over non-consecutive timestamps. We formalize WCCS by generalizing the classical $(α,β)$-core to a higher-order $(α,β,τ)$-wedge core, and by proposing a novel temporal wedge conductance metric that explicitly balances internal density and external sparsity. To solve WCCS efficiently, we first develop an online priority-driven filter-and-expand framework with several effective pruning techniques and a powerful geometric slope optimization for rapid temporal wedge conductance calculation. Subsequently, to further improve scalability, we propose an offline compressed index to accelerate search. Finally, comprehensive experiments on seven real-world datasets demonstrate the effectiveness, efficiency, and scalability of our solutions compared to eight competitors.

cs.SI↗

arXiv2021

Bi-Granularity Contrastive Learning for Post-Training in Few-Shot Scene

The major paradigm of applying a pre-trained language model to downstream tasks is to fine-tune it on labeled task data, which often suffers instability and low performance when the labeled examples are scarce.~One way to alleviate this problem is to apply post-training on unlabeled task data before fine-tuning, adapting the pre-trained model to target domains by contrastive learning that considers either token-level or sequence-level similarity. Inspired by the success of sequence masking, we argue that both token-level and sequence-level similarities can be captured with a pair of masked sequences.~Therefore, we propose complementary random masking (CRM) to generate a pair of masked sequences from an input sequence for sequence-level contrastive learning and then develop contrastive masked language modeling (CMLM) for post-training to integrate both token-level and sequence-level contrastive learnings.~Empirical results show that CMLM surpasses several recent post-training methods in few-shot settings without the need for data augmentation.

cs.CL↗

arXiv2020

Enhancing autonomy transparency: an option-centric rationale approach

While the advances in artificial intelligence and machine learning empower a new generation of autonomous systems for assisting human performance, one major concern arises from the human factors perspective: Humans have difficulty deciphering autonomy-generated solutions and increasingly perceive autonomy as a mysterious black box. The lack of transparency contributes to the lack of trust in autonomy and sub-optimal team performance. To enhance autonomy transparency, this study proposed an option-centric rationale display and evaluated its effectiveness. We developed a game Treasure Hunter wherein a human uncovers a map for treasures with the help from an intelligent assistant, and conducted a human-in-the-loop experiment with 34 participants. Results indicated that by conveying the intelligent assistant's decision-making rationale via the option-centric rationale display, participants had higher trust in the system and calibrated their trust faster. Additionally, higher trust led to higher acceptance of recommendations from the intelligent assistant, and in turn higher task performance.

cs.HC↗

arXiv2020

A Workload Adaptive Haptic Shared Control Scheme for Semi-Autonomous Driving

Haptic shared control is used to manage the control authority allocation between a human and an autonomous agent in semi-autonomous driving. Existing haptic shared control schemes, however, do not take full consideration of the human agent. To fill this research gap, this study presents a haptic shared control scheme that adapts to a human operator's workload, eyes on road and input torque in real-time. We conducted human-in-the-loop experiments with 24 participants. In the experiment, a human operator and an autonomy module for navigation shared the control of a simulated notional High Mobility Multipurpose Wheeled Vehicle (HMMWV) at a fixed speed. At the same time, the human operator performed a target detection task for surveillance. The autonomy could be either adaptive or non-adaptive to the above-mentioned human factors. Results indicate that the adaptive haptic control scheme resulted in significantly lower workload, higher trust in autonomy, better driving task performance and smaller control effort.

cs.HC↗

arXiv2013

Multi-Task Regularization with Covariance Dictionary for Linear Classifiers

In this paper we propose a multi-task linear classifier learning problem called D-SVM (Dictionary SVM). D-SVM uses a dictionary of parameter covariance shared by all tasks to do multi-task knowledge transfer among different tasks. We formally define the learning problem of D-SVM and show two interpretations of this problem, from both the probabilistic and kernel perspectives. From the probabilistic perspective, we show that our learning formulation is actually a MAP estimation on all optimization variables. We also show its equivalence to a multiple kernel learning problem in which one is trying to find a re-weighting kernel for features from a dictionary of basis (despite the fact that only linear classifiers are learned). Finally, we describe an alternative optimization scheme to minimize the objective function and present empirical studies to valid our algorithm.

cs.LG↗