arXiv subjects

Yiheng Liang

Publications and source records attributed to Yiheng Liang.

3 recordsLinked to original sources

arXiv2026

Replication Studies: Not Just a Copy

Replication studies revisit previous experimental findings for multiple purposes, including assessing reliability, exploring generalisability, and evaluating research methods. Such studies entail design decisions that a single replication label cannot fully capture, making their designs difficult to describe and compare. We present REPVIS2, a validated design space that describes how a replication study relates to a reference study across eight practical dimensions, each coded as identical, similar, or different. We refined our initial design space, REPVIS1, by applying it to a corpus of replication studies and validated REPVIS2 against a separate corpus. We then characterised 86 replication studies from 51 papers in visualisation through paired reading of the replication and reference reports. Studies often retained the task while changing the procedure, interface, environment, participant population, evidence, or analysis. Additions beyond the replication core were also common. REPVIS2 makes replication design explicit for characterisation, reporting, and planning. An interactive visualisation and supplementary materials are available at https://replication-study.github.io/.

cs.HC↗

arXiv2026

SCOPE: Signal-Calibrated On-Policy Distillation Enhancement with Dual-Path Adaptive Weighting

On-policy reinforcement learning has become the dominant paradigm for reasoning alignment in large language models, yet its sparse, outcome-level rewards make token-level credit assignment notoriously difficult. On-Policy Distillation (OPD) alleviates this by introducing dense, token-level KL supervision from a teacher model, but typically applies this supervision uniformly across all rollouts, ignoring fundamental differences in signal quality. We propose Signal-Calibrated On-Policy Distillation Enhancement (SCOPE), a dual-path adaptive training framework that routes on-policy rollouts by correctness into two complementary supervision paths. For incorrect trajectories, SCOPE performs teacher-perplexity-weighted KL distillation to prioritize instances where the teacher demonstrates genuine corrective capability, while down-weighting unreliable guidance. For correct trajectories, it applies student-perplexity-weighted MLE to concentrate reinforcement on low-confidence samples at the capability boundary rather than over-reinforcing already mastered ones. Both paths employ a group-level normalization to adaptively calibrate weight distributions, accounting for the intrinsic difficulty variance across prompts. Extensive experiments on six reasoning benchmarks show that SCOPE achieves an average relative improvement of 11.42% in Avg@32 and 7.30% in Pass@32 over competitive baselines, demonstrating its consistent effectiveness.

cs.LG↗

arXiv2026

Beyond Advocacy: A Design Space for Replication-Related Studies

The importance of replication is often discussed and advocated -- not only in the domains of visualization and HCI, but in all scientific areas. When replicating a study, design decisions need to be made with regards which aspects of the original study will remain the same and which will be altered. We present a supporting multi-dimensional design space framework within which such decisions can be identified, categorized, compared and analyzed. The framework treats replication experimental design as a pairwise comparison problem, and represents the design by four practical dimensions defined by three comparison levels. The design space is therefore a framework that can be used for both retrospective characterization and prospective planning. We provide worked examples, and relate our framework to other attempts at describing the scope of replication studies.

cs.HC↗