Institution
National Tsing Hua University
Recent research
- AI & ComputingOpen access
Anytime-Valid Stopping for Self-Consistency on Open Answer Sets
Self-consistency samples reasoning chains and returns the most frequent answer, and adaptive variants stop sampling early to save compute. Deployments act on the stopped answer, yet these rules check a posterior or window criterion as draws arrive, so the chance of stopping on an...
- Society & EconomicsOpen access
Economies of Scale to Consumption in Collective Households
ABSTRACT Browning et al. (2013) model collective‐household economies of scale in goods consumption by having each good be partly shared, instead of each good being public or private. We modify their model to achieve simple point identification and estimation of the economies of s...
- Society & EconomicsOpen access
Adaptive testing is entering language-model evaluation as an efficiency device, and its low item exposure is offered as incidental contamination protection (Li et al., 2026b). That argument assumes untargeted leakage, yet a maximum-information test asks predictable items that any...
- Health & MedicineOpen access
Background and objective: Telemedicine assesses voice from recordings sent over telephone and video links, where lossy speech codecs commonly sit between the voice and the screening model, and clinicians act on each patient’s flag. Codec studies report aggregate accuracy, yet two...
- AI & ComputingOpen access
Tabular foundation models answer every query about a table by in-context inference, so whether their answers to related queries fit one joint distribution is a direct test of whether they can be used as probabilistic models of the table. A recent test of marginalization consisten...
- Society & EconomicsOpen access
Selecting GPTQ Damping from the Calibration Set by Nested-Ridge Leave-One-Out
GPTQ-style quantization damps the calibration Hessian by a fixed fraction of its mean diagonal (one percent in the GPTQ authors' code, five percent in a maintained implementation), a choice that ignores how many calibration rows there are. Small tabular, time-series and text-embe...
ABSTRACT This work demonstrates a sustainable upcycling strategy that transforms discarded mask straps into flexible direct‐current triboelectric nanogenerators (DC‐TENGs) through an eco‐friendly NaCl‐assisted modification. The PEDOT:PSS/PVA/Pullulan composite simultaneously serv...
- AI & ComputingOpen access
Covariate Copies Count as Evidence in Time-Series Foundation Models
Covariate-aware time-series foundation models forecast a target from the related series supplied with it, and which series they receive depends on how a pipeline was built: the same sensor exported twice, a temperature in two units, one weather station joined in for several zones...
- Physics & SpaceOpen access
Abstract RR Lyrae stars (RRLs) are valuable probes of both Milky Way assembly and stellar-evolution physics. Using a sample 6240 RRLs obtained in the first year of the Dark Energy Spectroscopic Instrument (DESI) survey, we investigate the metallicity of RRLs and its correlation w...
- Physics & SpaceOpen access
Abstract We present the catalog of RR Lyrae stars (RRLs) observed in the first year of operations of the Dark Energy Spectroscopic Instrument (DESI) survey. This catalog contains 6240 RRLs out to ∼120 kpc from the Galactic center and over 12,000 individual epochs with homogeneous...
- Materials & EnergyOpen access
Abstract Novel [3+2] annulations between imines and 1,3-diacyl gold carbenes have been developed to provide 2-acyl-3-aminopyrrole derivatives efficiently. Mechanistic studies suggest that the reaction is initiated by nucleophilic addition of the imine to the 1,3-diacyl gold carbe...
- BiologyOpen access
Vertebrate organs undergo massive growth during the post-embryonic period. Yet, our understanding of how this organ-wide process is organized at single-cell resolution has been limited by an inability to monitor individual cells of different types throughout intact adult organs....
- Climate & Environment
Series‑connected fracture control on solubility‑Limited decay chain transport
Colloid-facilitated Th-230–Ra-226 decay-pair transport in series-connected fractured rock is examined as a conceptual mechanistic study using a GoldSim triple-continuum framework for fracture, skin, and matrix domains. Three fracture-segment types with distinct wetted perimeters...
- AI & ComputingOpen access
Tabular foundation models answer every query about a table by in-context inference, so whether their answers to related queries fit one joint distribution is a direct test of whether they can be used as probabilistic models of the table. A recent test of marginalization consisten...
- Climate & EnvironmentOpen access
Covariate Copies Count as Evidence in Time-Series Foundation Models
Covariate-aware time-series foundation models forecast a target from a group of related series, and the covariates supplied to them can be redundant: the same sensor exported twice, a temperature in two units, several nearby weather stations. How these models count such redundanc...
- Society & Economics
This study examined how highly trained open- and closed-skill athletes recruit neural resources during cue-supported motor imagery (MI) under two experimentally cued contexts: peak performance imagery (PPI) and routine training imagery (RTI). Thirty-four athletes (17 per group) c...
- Health & MedicineOpen access
Selecting GPTQ Damping from the Calibration Set by Nested-Ridge Leave-One-Out
GPTQ-style post-training quantization adds λI to the calibration Hessian before inverting it, with λ a fixed fraction of the mean diagonal (one percent in the GPTQ authors' code, five percent in a maintained implementation) that ignores how many calibration rows there are. Small...
- AI & ComputingOpen access
What pass@k Measures When Attempts Share State: Count Runs, Not Attempts
Search trees, planners and self-correcting agents make the attempts of one run share a random state, yet their results are often scored with a pass@k estimator built for independent draws, which public harnesses apply to whatever attempts they receive. Prior work separates the pr...
- Society & EconomicsOpen access
Practitioners choose between released 4-bit conversions of a model from published comparison sentences such as "AWQ consistently outperforms ... GPTQ ... (7B-70B)", although the files they serve, the serving stack and the way quality is measured all differ from what the sentence...
- Engineering & TechnologyOpen access
Controlled evidence on 4-bit weight quantization and in-context interference is proactive-only, limited to one 4-bit format and behavioural. In 1.5–8B open-weight models we cross proactive and retroactive layouts, with a recency-counterbalanced control, with interference dose, li...