arXiv subjects

Andrew Li

Publications and source records attributed to Andrew Li.

At least 19 recordsLinked to original sources

arXiv2026

Design and Modeling of the Charge Readout of a SiMOS Quantum Dot with a Single Electron Transistor and CryoCMOS

Single electron spin qubits trapped in SiMOS quantum dots are a promising technology for scaling to thousand- or million-qubit systems due to their compatibility with mature CMOS manufacturing processes. A readout system that combines a single electron transistor with a custom cryogenic CMOS amplification and digitization chain offers key advantages by avoiding the use of bulky RF components or room-temperature interconnects. We present design techniques and simulation results for an optimized qubit-SET cryoCMOS interface, culminating in the design of the QNDR1 ASIC, the first cryogenic readout ASIC designed under the Quandarum project, which targets the development of a many-channel spin qubit based detector for use in high energy physics.

cond-mat.mes-hall↗

arXiv2026

A Classification of Translation-Invariant Quantum Codes in Any Dimension

Quantum error-correcting codes with two-dimensional translation invariance are known to be equivalent to copies of the two-dimensional toric code. Such a simple classification is not possible for quantum codes with higher dimensional translation invariance due to the existence of multiple types of toric codes and infinite families of fracton codes. Here, we focus on D-dimensional translation-invariant quantum codes based on length-D chain complexes. This includes multivariate multicycle codes where the number of variables equals the number of cycles. We show that such codes are equivalent to copies of D-dimensional toric codes. This directly generalizes the classification result for two-dimensional translation-invariant codes.

quant-ph↗

arXiv2026

Beyond Static Leaderboards: Predictive Validity for the Evaluation of LLM Agents

Agent benchmarks are growing fast, but no single benchmark touches more than four or five of the dimensions that deployment exposes. This paper aggregates the largest coordinated deep-dive of one MCP-based industrial-agent benchmark to date: fourteen parallel implementation studies covering new asset classes (including a multi-modal visual extension), alternative orchestrations, retrieval strategies, reasoning modes, infrastructure optimizations, and evaluation-methodology probes. Consolidating those studies with seven prior agent benchmarks, we argue that aggregate-score leaderboards systematically underspecify deployed-agent evaluation. Rankings derived from aggregate scores do not transfer to out-of-distribution settings; recent public-to-hidden competition retrospectives provide direct empirical evidence of this rank instability. We propose ranking configurations by predictive validity, the correlation between in-sample and out-of-sample rank, rather than in-sample mean, and report a twelve-tier measurement apparatus that exposes the deployment-relevant dimensions HELM and its agent-era successors collapse. The position is operationalized through three falsifiable out-of-distribution criteria with explicit thresholds; existing evidence partly supports it but is too thin to confirm. We close with a pre-registered pilot design and a field-level vision for what the next generation of agentic benchmarks should report.

cs.AI↗

arXiv2026

cantnlp@DravidianLangTech 2026: organic domain adaptation improves multi-class hope speech detection in Tulu

This paper presents our systems and results for the Hope Speech Detection in Code-Mixed Tulu Language shared task at the Sixth Workshop on Speech, Vision, and Language Technologies for Dravidian Languages (DravidianLangTech-2026). We trained an XLM-RoBERTa-based text classification system for detecting hope speech in code-mixed Tulu social media comments. We compared this organically adapted hope speech detection model with our baseline model. On the development set, the organically adapted model outperformed the baseline system. While our submitted systems performed more modestly on the official test set, these results suggest that further adapting XLM-RoBERTa on organically collected Tulu social media text containing code-mixed and mixed-script variation can improve hope speech detection in code-mixed Tulu.

cs.CL↗

arXiv2025

Combinatorics of positional colored compositions

We consider colored compositions where only some parts are allowed different colors, depending on their locations in the composition. The counting sequences are obtained through generating functions. Connections to many other combinatorial objects are discussed, with combinatorial arguments provided and generalized for these observations.

math.CO↗

arXiv2025

How to Sell High-Dimensional Data Optimally

Motivated by the problem of selling large, proprietary data, we consider an information pricing problem proposed by Bergemann et al. that involves a decision-making buyer and a monopolistic seller. The seller has access to the underlying state of the world that determines the utility of the various actions the buyer may take. Since the buyer gains greater utility through better decisions resulting from more accurate assessments of the state, the seller can therefore promise the buyer supplemental information at a price. To contend with the fact that the seller may not be perfectly informed about the buyer's private preferences (or utility), we frame the problem of designing a data product as one where the seller designs a revenue-maximizing menu of statistical experiments. Prior work by Cai et al. showed that an optimal menu can be found in time polynomial in the state space, whereas we observe that the state space is naturally exponential in the dimension of the data. We propose an algorithm which, given only sampling access to the state space, provably generates a near-optimal menu with a number of samples independent of the state space. We then analyze a special case of high-dimensional Gaussian data, showing that (a) it suffices to consider scalar Gaussian experiments, (b) the optimal menu of such experiments can be found efficiently via a semidefinite program, and (c) full surplus extraction occurs if and only if a natural separation condition holds on the set of potential preferences of the buyer.

cs.GT↗

arXiv2025

Collective Communication Profiling of Modern-day Machine Learning Workloads

Machine Learning jobs, carried out on large number of distributed high performance systems, involve periodic communication using operations like AllReduce, AllGather, and Broadcast. These operations may create high bandwidth and bursty traffic patterns, leading to network congestion and packet loss, thus impacting the performance of these jobs. Hence it is imperative to analyze these patterns, which can be helpful in provisioning network resources depending on the type of machine learning workloads. In this poster we carry out extensive analysis of the collective communication behavior seen in a wide variety of models (ex. DeepSeek, GPT, Llama, etc.) To achieve this we instrument Nvidia Collective Communication Library logging functionality for richer context about the collectives and workloads. We adjust configuration parameters that influence collective communication behavior, such as parallelism, number of nodes, and model type. This overview presents and discusses some of the results on the collective communication behavior for the open source DeepSeek V3 inferencing model, which includes operation type and count, transfer sizes per operation, and request size distribution. Our analysis shows that it makes sense to rethink current collective communication frameworks and network topologies so as to accommodate the effect of network anomalies on the mentioned workloads.

cs.DC↗

arXiv2025

How Well Can General Vision-Language Models Learn Medicine By Watching Public Educational Videos?

Publicly available biomedical videos, such as those on YouTube, serve as valuable educational resources for medical students. Unlike standard machine learning datasets, these videos are designed for human learners, often mixing medical imagery with narration, explanatory diagrams, and contextual framing. In this work, we investigate whether such pedagogically rich, yet non-standardized and heterogeneous videos can effectively teach general-domain vision-language models biomedical knowledge. To this end, we introduce OpenBiomedVi, a biomedical video instruction tuning dataset comprising 1031 hours of video-caption and Q/A pairs, curated through a multi-step human-in-the-loop pipeline. Diverse biomedical video datasets are rare, and OpenBiomedVid fills an important gap by providing instruction-style supervision grounded in real-world educational content. Surprisingly, despite the informal and heterogeneous nature of these videos, the fine-tuned Qwen-2-VL models exhibit substantial performance improvements across most benchmarks. The 2B model achieves gains of 98.7% on video tasks, 71.2% on image tasks, and 0.2% on text tasks. The 7B model shows improvements of 37.09% on video and 11.2% on image tasks, with a slight degradation of 2.7% on text tasks compared to their respective base models. To address the lack of standardized biomedical video evaluation datasets, we also introduce two new expert curated benchmarks, MIMICEchoQA and SurgeryVideoQA. On these benchmarks, the 2B model achieves gains of 99.1% and 98.1%, while the 7B model shows gains of 22.5% and 52.1%, respectively, demonstrating the models' ability to generalize and perform biomedical video understanding on cleaner and more standardized datasets than those seen during training. These results suggest that educational videos created for human learning offer a surprisingly effective training signal for biomedical VLMs.

cs.CV↗

arXiv2025

cantnlp@DravidianLangTech2025: A Bag-of-Sounds Approach to Multimodal Hate Speech Detection

This paper presents the systems and results for the Multimodal Social Media Data Analysis in Dravidian Languages (MSMDA-DL) shared task at the Fifth Workshop on Speech, Vision, and Language Technologies for Dravidian Languages (DravidianLangTech-2025). We took a `bag-of-sounds' approach by training our hate speech detection system on the speech (audio) data using transformed Mel spectrogram measures. While our candidate model performed poorly on the test set, our approach offered promising results during training and development for Malayalam and Tamil. With sufficient and well-balanced training data, our results show that it is feasible to use both text and speech (audio) data in the development of multimodal hate speech detection systems.

cs.CL↗

arXiv2025

Practical limitations of the switching theorem for adiabatic state preparation

The viability of adiabatic quantum computation depends on the slow evolution of the Hamiltonian. The adiabatic switching theorem provides an asymptotic series for error estimates in $1/T$, based on the lowest non-zero derivative of the Hamiltonian and its eigenvalues at the endpoints. Modifications at the endpoints in practical implementations can modify this scaling behavior, suggesting opportunities for error reduction by altering endpoint behavior while keeping intermediate evolution largely unchanged. Such modifications can significantly reduce errors for long evolution times, but they may also require exceedingly long timescales to reach the hyperadiabatic regime, limiting their practicality. This paper explores the transition between the adiabatic and hyperadiabatic regimes in simple low-dimensional Hamiltonians, highlighting the impact of modifications of the endpoints on approaching the asymptotic behavior described by the switching theorem.

quant-ph↗

arXiv2025

SMIR: Efficient Synthetic Data Pipeline To Improve Multi-Image Reasoning

Vision-Language Models (VLMs) excel at understanding single images, aided by high-quality instruction datasets. However, multi-image reasoning remains underexplored in the open-source community due to two key challenges: (1) scaling datasets with correlated images and complex reasoning instructions is resource-intensive, and (2) robust evaluation benchmarks for multi-image tasks are lacking. To address this, we introduce SMiR, a synthetic data-generation pipeline for multi-image reasoning, along with a high-quality dataset generated using this pipeline. SMiR efficiently extracts correlated images via multimodal embeddings, integrates visual and descriptive information, and leverages open-source LLMs to generate quality instructions. Using this approach, we produce 160K synthetic training samples, offering a cost-effective alternative to closed-source solutions. Additionally, we present SMiR-Bench, a multi-image reasoning benchmark comprising 200 diverse examples across seven complex reasoning tasks. SMiR-Bench is multi-turn and employs a VLM judge to evaluate free-form responses, providing a comprehensive assessment of model expressiveness and reasoning capability across modalities. We demonstrate the effectiveness of SMiR by fine-tuning open-source VLMs and evaluating them on SMiR-Bench.

cs.CV↗

arXiv2024

$S^5$: New insights from deep spectroscopic observations of the tidal tails of the globular clusters NGC 1261 and NGC 1904

As globular clusters (GCs) orbit the Milky Way, their stars are tidally stripped forming tidal tails that follow the orbit of the clusters around the Galaxy. The morphology of these tails is complex and shows correlations with the phase of the orbit and the orbital angular velocity, especially for GCs on eccentric orbits. Here, we focus on two GCs, NGC 1261 and NGC 1904, that have potentially been accreted alongside Gaia-Enceladus and that have shown signatures of having, in addition of tidal tails, structures formed by distributions of extra-tidal stars that are misaligned with the general direction of the clusters' respective orbits. To provide an explanation for the formation of these structures, we make use of spectroscopic measurements from the Southern Stellar Stream Spectroscopic Survey ($S^5$) as well as proper motion measurements from Gaia's third data release (DR3), and apply a Bayesian mixture modeling approach to isolate high-probability member stars. We recover extra-tidal features similar to those found in Shipp et al. (2018) surrounding each cluster. We conduct N-body simulations and compare the expected distribution and variation in the dynamical parameters along the orbit with those of our potential member sample. Furthermore, we use Dark Energy Camera (DECam) photometry to inspect the distribution of the member stars in the color-magnitude diagram (CMD). We find that the potential members agree reasonably with the N-body simulations and that the majority of them follow a simple stellar population-like distribution in the CMD which is characteristic of GCs. In the case of NGC 1904, we clearly detect the tidal debris escaping the inner and outer Lagrange points which are expected to be prominent when at or close to the apocenter of its orbit. Our analysis allows for further exploration of other GCs in the Milky Way that exhibit similar extra-tidal features.

astro-ph.GA↗

arXiv2024

EmbedLLM: Learning Compact Representations of Large Language Models

With hundreds of thousands of language models available on Huggingface today, efficiently evaluating and utilizing these models across various downstream, tasks has become increasingly critical. Many existing methods repeatedly learn task-specific representations of Large Language Models (LLMs), which leads to inefficiencies in both time and computational resources. To address this, we propose EmbedLLM, a framework designed to learn compact vector representations, of LLMs that facilitate downstream applications involving many models, such as model routing. We introduce an encoder-decoder approach for learning such embeddings, along with a systematic framework to evaluate their effectiveness. Empirical results show that EmbedLLM outperforms prior methods in model routing both in accuracy and latency. Additionally, we demonstrate that our method can forecast a model's performance on multiple benchmarks, without incurring additional inference cost. Extensive probing experiments validate that the learned embeddings capture key model characteristics, e.g. whether the model is specialized for coding tasks, even without being explicitly trained on them. We open source our dataset, code and embedder to facilitate further research and application.

cs.CL↗

arXiv2024

Incremental Comprehension of Garden-Path Sentences by Large Language Models: Semantic Interpretation, Syntactic Re-Analysis, and Attention

When reading temporarily ambiguous garden-path sentences, misinterpretations sometimes linger past the point of disambiguation. This phenomenon has traditionally been studied in psycholinguistic experiments using online measures such as reading times and offline measures such as comprehension questions. Here, we investigate the processing of garden-path sentences and the fate of lingering misinterpretations using four large language models (LLMs): GPT-2, LLaMA-2, Flan-T5, and RoBERTa. The overall goal is to evaluate whether humans and LLMs are aligned in their processing of garden-path sentences and in the lingering misinterpretations past the point of disambiguation, especially when extra-syntactic information (e.g., a comma delimiting a clause boundary) is present to guide processing. We address this goal using 24 garden-path sentences that have optional transitive and reflexive verbs leading to temporary ambiguities. For each sentence, there are a pair of comprehension questions corresponding to the misinterpretation and the correct interpretation. In three experiments, we (1) measure the dynamic semantic interpretations of LLMs using the question-answering task; (2) track whether these models shift their implicit parse tree at the point of disambiguation (or by the end of the sentence); and (3) visualize the model components that attend to disambiguating information when processing the question probes. These experiments show promising alignment between humans and LLMs in the processing of garden-path sentences, especially when extra-syntactic information is available to guide processing.

cs.CL↗

arXiv2024

Color code decoder with improved scaling for correcting circuit-level noise

Two-dimensional color codes are a promising candidate for fault-tolerant quantum computing, as they have high encoding rates, transversal implementation of logical Clifford gates, and resource-efficient magic state preparation schemes. However, decoding color codes presents a significant challenge due to their structure, where elementary errors violate three checks instead of just two (a key feature in surface code decoding), and the complexity of extracting syndrome is greater. We introduce an efficient color-code decoder that tackles these issues by combining two matching decoders for each color, generalized to handle circuit-level noise by employing detector error models. We provide comprehensive analyses of the decoder, covering its threshold and sub-threshold scaling both for bit-flip noise with ideal measurements and for circuit-level noise. Our simulations reveal that this decoding strategy nearly reaches the best possible scaling of logical failure ($p_\mathrm{fail} \sim p^{d/2}$) for both noise models, where $p$ is the noise strength, in the regime of interest for fault-tolerant quantum computing. While its noise thresholds are comparable with other matching-based decoders for color codes ($8.2\%$ for bit-flip noise and $0.46\%$ for circuit-level noise), the scaling of logical failure rates below threshold significantly outperforms the best matching-based decoders.

quant-ph↗

arXiv2024

Deep Few-view High-resolution Photon-counting CT at Halved Dose for Extremity Imaging

X-ray photon-counting computed tomography (PCCT) for extremity allows multi-energy high-resolution (HR) imaging but its radiation dose can be further improved. Despite the great potential of deep learning techniques, their application in HR volumetric PCCT reconstruction has been challenged by the large memory burden, training data scarcity, and domain gap issues. In this paper, we propose a deep learning-based approach for PCCT image reconstruction at halved dose and doubled speed validated in a New Zealand clinical trial. Specifically, we design a patch-based volumetric refinement network to alleviate the GPU memory limitation, train network with synthetic data, and use model-based iterative refinement to bridge the gap between synthetic and clinical data. Our results in a reader study of 8 patients from the clinical trial demonstrate a great potential to cut the radiation dose to half that of the clinical PCCT standard without compromising image quality and diagnostic value.

physics.med-ph↗

arXiv2024

On the set partitions that require maximum sorts through the $aba-$avoiding stack

Recently, Xia introduced a deterministic variation $\phi_{\sigma}$ of Defant and Kravitz's stack-sorting maps for set partitions and showed that any set partition $p$ is sorted by $\phi^{N(p)}_{aba}$, where $N(p)$ is the number of distinct alphabets in $p$. Xia then asked which set partitions $p$ are not sorted by $\phi_{aba}^{N(p)-1}$. In this note, we prove that the minimal length of a set partition $p$ that is not sorted by $\phi_{aba}^{N(p)-1}$ is $2N(p)$. Then we show that there is only one set partition of length $2N(p)$ and ${{N(p) + 1} \choose 2} + 2{N(p) \choose 2}$ set partitions of length $2N(p)+1$ that are not sorted by $\phi_{aba}^{N(p)-1}$.

math.CO↗

arXiv2023

Markdown Pricing Under an Unknown Parametric Demand Model

Consider a single-product revenue-maximization problem where the seller monotonically decreases the price in $n$ rounds with an unknown demand model coming from a given family. Without monotonicity, the minimax regret is $\tilde O(n^{2/3})$ for the Lipschitz demand family and $\tilde O(n^{1/2})$ for a general class of parametric demand models. With monotonicity, the minimax regret is $\tilde O(n^{3/4})$ if the revenue function is Lipschitz and unimodal. However, the minimax regret for parametric families remained open. In this work, we provide a complete settlement for this fundamental problem. We introduce the crossing number to measure the complexity of a family of demand functions. In particular, the family of degree-$k$ polynomials has a crossing number $k$. Based on conservatism under uncertainty, we present (i) a policy with an optimal $\Theta(\log^2 n)$ regret for families with crossing number $k=0$, and (ii) another policy with an optimal $\tilde \Theta(n^{k/(k+1)})$ regret when $k\ge 1$. These bounds are asymptotically higher than the $\tilde O(\log n)$ and $\tilde \Theta(\sqrt n)$ minimax regret for the same families without the monotonicity constraint.

cs.LG↗