Showing 1–50 of 71 results for author: Adams, G

The Chern-Simons Natural Boundary and Black Hole Entropy

Authors: Griffen Adams, Gerald V. Dunne

Abstract: The method of resurgent continuation of transseries reveals a new correspondence between the $q$-series for enumerating degeneracies of quarter-BPS states in supersymmetric black holes and $\hat{Z}$ invariants of Chern-Simons theory on a class of 3 dimensional orientation-reversed manifolds. The method of resurgent continuation of transseries reveals a new correspondence between the $q$-series for enumerating degeneracies of quarter-BPS states in supersymmetric black holes and $\hat{Z}$ invariants of Chern-Simons theory on a class of 3 dimensional orientation-reversed manifolds. △ Less

Submitted 4 March, 2026; originally announced March 2026.

Comments: 24 pages, 4 figures

Plug to Place: Indoor Multimedia Geolocation from Electrical Sockets for Digital Investigation

Authors: Kanwal Aftab, Graham Adams, Mark Scanlon

Abstract: Computer vision is a rapidly evolving field, giving rise to powerful new tools and techniques in digital forensic investigation, and shows great promise for novel digital forensic applications. One such application, indoor multimedia geolocation, has the potential to become a crucial aid for law enforcement in the fight against human trafficking, child exploitation, and other serious crimes. While… ▽ More Computer vision is a rapidly evolving field, giving rise to powerful new tools and techniques in digital forensic investigation, and shows great promise for novel digital forensic applications. One such application, indoor multimedia geolocation, has the potential to become a crucial aid for law enforcement in the fight against human trafficking, child exploitation, and other serious crimes. While outdoor multimedia geolocation has been widely explored, its indoor counterpart remains underdeveloped due to challenges such as similar room layouts, frequent renovations, visual ambiguity, indoor lighting variability, unreliable GPS signals, and limited datasets in sensitive domains. This paper introduces a pipeline that uses electric sockets as consistent indoor markers for geolocation, since plug socket types are standardised by country or region. The three-stage deep learning pipeline detects plug sockets (YOLOv11, mAP@0.5 = 0.843), classifies them into one of 12 plug socket types (Xception, accuracy = 0.912), and maps the detected socket types to countries (accuracy = 0.96 at >90% threshold confidence). To address data scarcity, two dedicated datasets were created: socket detection dataset of 2,328 annotated images expanded to 4,072 through augmentation, and a classification dataset of 3,187 images across 12 plug socket classes. The pipeline was evaluated on the Hotels-50K dataset, focusing on the TraffickCam subset of crowd-sourced hotel images, which capture real-world conditions such as poor lighting and amateur angles. This dataset provides a more realistic evaluation than using professional, well-lit, often wide-angle images from travel websites. This framework demonstrates a practical step toward real-world digital forensic applications. The code, trained models, and the data for this paper are available open source. △ Less

Submitted 18 December, 2025; originally announced December 2025.

$c_{\rm eff}$ from Resurgence at the Stokes Line

Authors: Griffen Adams, Ovidiu Costin, Gerald V. Dunne, Sergei Gukov, Oğuz Öner

Abstract: In recent papers [1,2], a new method to cross the natural boundary has been proposed, and applied to Mordell-Borel integrals arising in the study of Chern-Simons theory, based on decompositions into {\it resurgent cyclic orbits}. Resurgent analysis on the Stokes line leads to a unique transseries decomposition in terms of unary false theta functions, which can be continued across the natural bound… ▽ More In recent papers [1,2], a new method to cross the natural boundary has been proposed, and applied to Mordell-Borel integrals arising in the study of Chern-Simons theory, based on decompositions into {\it resurgent cyclic orbits}. Resurgent analysis on the Stokes line leads to a unique transseries decomposition in terms of unary false theta functions, which can be continued across the natural boundary to produce dual $q$-series whose integer-valued coefficients enumerate BPS states. This constitutes a deeper new manifestation of resurgence in quantum field theoretic path integrals. In this paper we show that the algebraic structure of the {\it resurgent cyclic orbits}, combined with just the leading term of the $q$-series, completely determines the large order rate of growth of the dual $q$-series coefficients. The essential exponent of this asymptotic growth has a Cardy-like interpretation [10] of an effective central charge in a 3 dimensional quantum field theory with $\mathcal{N}=2$ supersymmetry related to the Chern-Simons theory through the $3d$-$3d$ correspondence. △ Less

Submitted 18 September, 2025; v1 submitted 13 August, 2025; originally announced August 2025.

Comments: 41 pages, 11 figures; v3 lists further comparisons with mock Jacobi forms

Report number: BONN-TH-2025-13, MPIM-Bonn-2024

Journal ref: Journal of High Energy Physics 2026-02-05

Orientation Reversal and the Chern-Simons Natural Boundary

Authors: Griffen Adams, Ovidiu Costin, Gerald V. Dunne, Sergei Gukov, Oğuz Öner

Abstract: We show that the fundamental property of preservation of relations, underlying resurgent analysis, provides a new perspective on crossing a natural boundary, an important general problem in theoretical and mathematical physics. This reveals a deeper rigidity of resurgence in a quantum field theory. We study the non-perturbative completion of complex Chern-Simons theory that associates to a 3-manif… ▽ More We show that the fundamental property of preservation of relations, underlying resurgent analysis, provides a new perspective on crossing a natural boundary, an important general problem in theoretical and mathematical physics. This reveals a deeper rigidity of resurgence in a quantum field theory. We study the non-perturbative completion of complex Chern-Simons theory that associates to a 3-manifold a collection of $q$-series invariants labeled by Spin$^c$ structures, for which crossing the natural boundary corresponds to orientation reversal of the 3-manifold. Our new resurgent perspective leads to a practical numerical algorithm that generates $q$-series which are dual to unary $q$-series composed of false theta functions. Until recently, these duals were only known in a limited number of cases, essentially based on Ramanujan's mock theta functions, and the common belief was that the more general duals might not even exist. Resurgence analysis identifies as primary objects Mordell integrals: transforms of resurgent functions. Their unique Borel summed transseries decomposition on either side of the Stokes line is the unique decomposition into real and imaginary parts. The latter are combinations of unary $q$-series in terms of $q$ and its modular counterpart $\tilde{q}$, and are resurgent by construction. The Mordell integral is analytic across the natural boundary of the $q$ and $\tilde{q}$ series, and uniqueness of a similar decomposition which preserves algebraic relations on the other side of the boundary defines the unique boundary crossing of the $q$ series. This continuation can be efficiently implemented numerically. This identifies known unique mock modular identities, and extends well beyond. The resurgent approach reveals new aspects, and is very different from other approaches based on indefinite theta series, Appell-Lerch sums, and logarithmic vertex operator algebras. △ Less

Submitted 12 June, 2025; v1 submitted 20 May, 2025; originally announced May 2025.

Comments: 63 pages, 13 figures, 20 tables; v2 ancillary files added

Report number: BONN-TH-2025-13, MPIM-Bonn-2024

Journal ref: JHEP 08 (2025) 154

Spreading of highly cohesive metal powders with transverse oscillation kinematics

Authors: Reimar Weissbach, Garrett Adams, Patrick M. Praegla, Christoph Meier, A. John Hart

Abstract: Powder bed additive manufacturing processes such as laser powder bed fusion (LPBF) or binder jetting (BJ) benefit from using fine (D50 $\leq20~μm$) powders. However, the increasing level of cohesion with decreasing particle size makes spreading a uniform and continuous layer challenging. As a result, LPBF typically employs a coarser size distribution, and rotating roller mechanisms are used in BJ… ▽ More Powder bed additive manufacturing processes such as laser powder bed fusion (LPBF) or binder jetting (BJ) benefit from using fine (D50 $\leq20~μm$) powders. However, the increasing level of cohesion with decreasing particle size makes spreading a uniform and continuous layer challenging. As a result, LPBF typically employs a coarser size distribution, and rotating roller mechanisms are used in BJ machines, that can create wave-like surface profiles due to roller run-out. In this work, a transverse oscillation kinematic for powder spreading is proposed, explored computationally, and validated experimentally. Simulations are performed using an integrated discrete element-finite element (DEM-FEM) framework and predict that transverse oscillation of a non-rotating roller facilitates the spreading of dense powder layers (beyond 50% packing fraction) with a high level of robustness to kinematic parameters. The experimental study utilizes a custom-built mechanized powder spreading testbed and X-ray transmission imaging for the analysis of spread powder layers. Experimental results generally validate the computational results, however, also exhibit parasitic layer cracking. For transverse oscillation frequencies above 200 Hz, powder layers of high packing fraction (between 50-60%) were formed, and for increased layer thicknesses, highly uniform and continuous layers were deposited. Statistical analysis of the experimental powder layer morphology as a function of kinematic spreading parameters revealed that an increasing transverse surface velocity improves layer uniformity and reduces cracking defects. This suggests that with minor improvements to the machine design, the proposed transverse oscillation kinematic has the potential to result in thin and consistently uniform powder layers of highly cohesive powder. △ Less

Submitted 26 April, 2025; originally announced April 2025.

Smarter, Better, Faster, Longer: A Modern Bidirectional Encoder for Fast, Memory Efficient, and Long Context Finetuning and Inference

Authors: Benjamin Warner, Antoine Chaffin, Benjamin Clavié, Orion Weller, Oskar Hallström, Said Taghadouini, Alexis Gallagher, Raja Biswas, Faisal Ladhak, Tom Aarsen, Nathan Cooper, Griffin Adams, Jeremy Howard, Iacopo Poli

Abstract: Encoder-only transformer models such as BERT offer a great performance-size tradeoff for retrieval and classification tasks with respect to larger decoder-only models. Despite being the workhorse of numerous production pipelines, there have been limited Pareto improvements to BERT since its release. In this paper, we introduce ModernBERT, bringing modern model optimizations to encoder-only models… ▽ More Encoder-only transformer models such as BERT offer a great performance-size tradeoff for retrieval and classification tasks with respect to larger decoder-only models. Despite being the workhorse of numerous production pipelines, there have been limited Pareto improvements to BERT since its release. In this paper, we introduce ModernBERT, bringing modern model optimizations to encoder-only models and representing a major Pareto improvement over older encoders. Trained on 2 trillion tokens with a native 8192 sequence length, ModernBERT models exhibit state-of-the-art results on a large pool of evaluations encompassing diverse classification tasks and both single and multi-vector retrieval on different domains (including code). In addition to strong downstream performance, ModernBERT is also the most speed and memory efficient encoder and is designed for inference on common GPUs. △ Less

Submitted 19 December, 2024; v1 submitted 18 December, 2024; originally announced December 2024.

Effects of Metallicity on Graphite, TiC, and SiC Condensation in Carbon Stars

Authors: Gabrielle Adams, Katharina Lodders

Abstract: From transmission electron microscopy and other laboratory studies of presolar grains, the implicit condensation sequence of carbon-bearing condensates in circumstellar envelopes of carbon stars is (from first to last) TiC-graphite-SiC. We use thermochemical equilibrium condensation calculations and show that the condensation sequence of TiC, graphite, and SiC depends on metallicity in addition to… ▽ More From transmission electron microscopy and other laboratory studies of presolar grains, the implicit condensation sequence of carbon-bearing condensates in circumstellar envelopes of carbon stars is (from first to last) TiC-graphite-SiC. We use thermochemical equilibrium condensation calculations and show that the condensation sequence of TiC, graphite, and SiC depends on metallicity in addition to C/O ratio and total pressure. Calculations were performed for a characteristic carbon star ratio of C/O = 1.2 from 1E-10 to 1E-4 bars total pressure and for uniform metallicity variations ranging from 0.01 to 100 times solar elemental abundances. TiC always condenses before SiC, and the carbide condensation temperatures increase with increasing metallicity and total pressure. Graphite, however, can condense in a cooling circumstellar envelope before TiC, between TiC and SiC, or after SiC, depending on the carbon-bearing gas chemistry, which is dependent on metallicity and total pressure. Analytical expressions for the graphite, TiC, and SiC condensation temperatures as functions of metallicity and total pressure are presented. The inferred sequence from laboratory presolar grain studies, TiC-graphite-SiC, is favored under equilibrium conditions at solar and subsolar metallicities between ~1E-5 to 1E-8 bar total pressure within circumstellar envelopes of carbon stars with nominal C/O = 1.2. We also explored the dependence of the sequence at C/O ratios of 1.1 and 3.0 and found that as the C/O ratio increases, the TiC-graphite-SiC region shifts towards higher total pressures and lower metallicities. △ Less

Submitted 3 June, 2025; v1 submitted 18 November, 2024; originally announced November 2024.

Comments: 33 pages, 9 figures, 2 tables. Accepted to The Astrophysical Journal, March 2025

Journal ref: ApJ 985 35 (2025)

Environment Scan of Generative AI Infrastructure for Clinical and Translational Science

Authors: Betina Idnay, Zihan Xu, William G. Adams, Mohammad Adibuzzaman, Nicholas R. Anderson, Neil Bahroos, Douglas S. Bell, Cody Bumgardner, Thomas Campion, Mario Castro, James J. Cimino, I. Glenn Cohen, David Dorr, Peter L Elkin, Jungwei W. Fan, Todd Ferris, David J. Foran, David Hanauer, Mike Hogarth, Kun Huang, Jayashree Kalpathy-Cramer, Manoj Kandpal, Niranjan S. Karnik, Avnish Katoch, Albert M. Lai , et al. (32 additional authors not shown)

Abstract: This study reports a comprehensive environmental scan of the generative AI (GenAI) infrastructure in the national network for clinical and translational science across 36 institutions supported by the Clinical and Translational Science Award (CTSA) Program led by the National Center for Advancing Translational Sciences (NCATS) of the National Institutes of Health (NIH) at the United States. With t… ▽ More This study reports a comprehensive environmental scan of the generative AI (GenAI) infrastructure in the national network for clinical and translational science across 36 institutions supported by the Clinical and Translational Science Award (CTSA) Program led by the National Center for Advancing Translational Sciences (NCATS) of the National Institutes of Health (NIH) at the United States. With the rapid advancement of GenAI technologies, including large language models (LLMs), healthcare institutions face unprecedented opportunities and challenges. This research explores the current status of GenAI integration, focusing on stakeholder roles, governance structures, and ethical considerations by administering a survey among leaders of health institutions (i.e., representing academic medical centers and health systems) to assess the institutional readiness and approach towards GenAI adoption. Key findings indicate a diverse range of institutional strategies, with most organizations in the experimental phase of GenAI deployment. The study highlights significant variations in governance models, with a strong preference for centralized decision-making but notable gaps in workforce training and ethical oversight. Moreover, the results underscore the need for a more coordinated approach to GenAI governance, emphasizing collaboration among senior leaders, clinicians, information technology staff, and researchers. Our analysis also reveals concerns regarding GenAI bias, data security, and stakeholder trust, which must be addressed to ensure the ethical and effective implementation of GenAI technologies. This study offers valuable insights into the challenges and opportunities of GenAI integration in healthcare, providing a roadmap for institutions aiming to leverage GenAI for improved quality of care and operational efficiency. △ Less

Submitted 27 September, 2024; originally announced October 2024.

Reducing the Footprint of Multi-Vector Retrieval with Minimal Performance Impact via Token Pooling

Authors: Benjamin Clavié, Antoine Chaffin, Griffin Adams

Abstract: Over the last few years, multi-vector retrieval methods, spearheaded by ColBERT, have become an increasingly popular approach to Neural IR. By storing representations at the token level rather than at the document level, these methods have demonstrated very strong retrieval performance, especially in out-of-domain settings. However, the storage and memory requirements necessary to store the large… ▽ More Over the last few years, multi-vector retrieval methods, spearheaded by ColBERT, have become an increasingly popular approach to Neural IR. By storing representations at the token level rather than at the document level, these methods have demonstrated very strong retrieval performance, especially in out-of-domain settings. However, the storage and memory requirements necessary to store the large number of associated vectors remain an important drawback, hindering practical adoption. In this paper, we introduce a simple clustering-based token pooling approach to aggressively reduce the number of vectors that need to be stored. This method can reduce the space & memory footprint of ColBERT indexes by 50% with virtually no retrieval performance degradation. This method also allows for further reductions, reducing the vector count by 66%-to-75% , with degradation remaining below 5% on a vast majority of datasets. Importantly, this approach requires no architectural change nor query-time processing, and can be used as a simple drop-in during indexation with any ColBERT-like model. △ Less

Submitted 22 September, 2024; originally announced September 2024.

STORYSUMM: Evaluating Faithfulness in Story Summarization

Authors: Melanie Subbiah, Faisal Ladhak, Akankshya Mishra, Griffin Adams, Lydia B. Chilton, Kathleen McKeown

Abstract: Human evaluation has been the gold standard for checking faithfulness in abstractive summarization. However, with a challenging source domain like narrative, multiple annotators can agree a summary is faithful, while missing details that are obvious errors only once pointed out. We therefore introduce a new dataset, STORYSUMM, comprising LLM summaries of short stories with localized faithfulness l… ▽ More Human evaluation has been the gold standard for checking faithfulness in abstractive summarization. However, with a challenging source domain like narrative, multiple annotators can agree a summary is faithful, while missing details that are obvious errors only once pointed out. We therefore introduce a new dataset, STORYSUMM, comprising LLM summaries of short stories with localized faithfulness labels and error explanations. This benchmark is for evaluation methods, testing whether a given method can detect challenging inconsistencies. Using this dataset, we first show that any one human annotation protocol is likely to miss inconsistencies, and we advocate for pursuing a range of methods when establishing ground truth for a summarization dataset. We finally test recent automatic metrics and find that none of them achieve more than 70% balanced accuracy on this task, demonstrating that it is a challenging benchmark for future work in faithfulness evaluation. △ Less

Submitted 1 April, 2025; v1 submitted 8 July, 2024; originally announced July 2024.

Comments: EMNLP Main 2024

Generating Faithful and Complete Hospital-Course Summaries from the Electronic Health Record

Authors: Griffin Adams

Abstract: The rapid adoption of Electronic Health Records (EHRs) has been instrumental in streamlining administrative tasks, increasing transparency, and enabling continuity of care across providers. An unintended consequence of the increased documentation burden, however, has been reduced face-time with patients and, concomitantly, a dramatic rise in clinician burnout. In this thesis, we pinpoint a particu… ▽ More The rapid adoption of Electronic Health Records (EHRs) has been instrumental in streamlining administrative tasks, increasing transparency, and enabling continuity of care across providers. An unintended consequence of the increased documentation burden, however, has been reduced face-time with patients and, concomitantly, a dramatic rise in clinician burnout. In this thesis, we pinpoint a particularly time-intensive, yet critical, documentation task: generating a summary of a patient's hospital admissions, and propose and evaluate automated solutions. In Chapter 2, we construct a dataset based on 109,000 hospitalizations (2M source notes) and perform exploratory analyses to motivate future work on modeling and evaluation [NAACL 2021]. In Chapter 3, we address faithfulness from a modeling perspective by revising noisy references [EMNLP 2022] and, to reduce the reliance on references, directly calibrating model outputs to metrics [ACL 2023]. These works relied heavily on automatic metrics as human annotations were limited. To fill this gap, in Chapter 4, we conduct a fine-grained expert annotation of system errors in order to meta-evaluate existing metrics and better understand task-specific issues of domain adaptation and source-summary alignments. To learn a metric less correlated to extractiveness (copy-and-paste), we derive noisy faithfulness labels from an ensemble of existing metrics and train a faithfulness classifier on these pseudo labels [MLHC 2023]. Finally, in Chapter 5, we demonstrate that fine-tuned LLMs (Mistral and Zephyr) are highly prone to entity hallucinations and cover fewer salient entities. We improve both coverage and faithfulness by performing sentence-level entity planning based on a set of pre-computed salient entities from the source text, which extends our work on entity-guided news summarization [ACL, 2023], [EMNLP, 2023]. △ Less

Submitted 1 April, 2024; originally announced April 2024.

Comments: PhD thesis

SPEER: Sentence-Level Planning of Long Clinical Summaries via Embedded Entity Retrieval

Authors: Griffin Adams, Jason Zucker, Noémie Elhadad

Abstract: Clinician must write a lengthy summary each time a patient is discharged from the hospital. This task is time-consuming due to the sheer number of unique clinical concepts covered in the admission. Identifying and covering salient entities is vital for the summary to be clinically useful. We fine-tune open-source LLMs (Mistral-7B-Instruct and Zephyr-7B-beta) on the task and find that they generate… ▽ More Clinician must write a lengthy summary each time a patient is discharged from the hospital. This task is time-consuming due to the sheer number of unique clinical concepts covered in the admission. Identifying and covering salient entities is vital for the summary to be clinically useful. We fine-tune open-source LLMs (Mistral-7B-Instruct and Zephyr-7B-beta) on the task and find that they generate incomplete and unfaithful summaries. To increase entity coverage, we train a smaller, encoder-only model to predict salient entities, which are treated as content-plans to guide the LLM. To encourage the LLM to focus on specific mentions in the source notes, we propose SPEER: Sentence-level Planning via Embedded Entity Retrieval. Specifically, we mark each salient entity span with special "{ }" boundary tags and instruct the LLM to retrieve marked spans before generating each sentence. Sentence-level planning acts as a form of state tracking in that the model is explicitly recording the entities it uses. We fine-tune Mistral and Zephyr variants on a large-scale, diverse dataset of ~167k in-patient hospital admissions and evaluate on 3 datasets. SPEER shows gains in both coverage and faithfulness metrics over non-guided and guided baselines. △ Less

Submitted 26 September, 2024; v1 submitted 4 January, 2024; originally announced January 2024.

Comments: COLM 2024

From Sparse to Dense: GPT-4 Summarization with Chain of Density Prompting

Authors: Griffin Adams, Alexander Fabbri, Faisal Ladhak, Eric Lehman, Noémie Elhadad

Abstract: Selecting the ``right'' amount of information to include in a summary is a difficult task. A good summary should be detailed and entity-centric without being overly dense and hard to follow. To better understand this tradeoff, we solicit increasingly dense GPT-4 summaries with what we refer to as a ``Chain of Density'' (CoD) prompt. Specifically, GPT-4 generates an initial entity-sparse summary be… ▽ More Selecting the ``right'' amount of information to include in a summary is a difficult task. A good summary should be detailed and entity-centric without being overly dense and hard to follow. To better understand this tradeoff, we solicit increasingly dense GPT-4 summaries with what we refer to as a ``Chain of Density'' (CoD) prompt. Specifically, GPT-4 generates an initial entity-sparse summary before iteratively incorporating missing salient entities without increasing the length. Summaries generated by CoD are more abstractive, exhibit more fusion, and have less of a lead bias than GPT-4 summaries generated by a vanilla prompt. We conduct a human preference study on 100 CNN DailyMail articles and find that that humans prefer GPT-4 summaries that are more dense than those generated by a vanilla prompt and almost as dense as human written summaries. Qualitative analysis supports the notion that there exists a tradeoff between informativeness and readability. 500 annotated CoD summaries, as well as an extra 5,000 unannotated summaries, are freely available on HuggingFace (https://huggingface.co/datasets/griffin/chain_of_density). △ Less

Submitted 8 September, 2023; originally announced September 2023.

Comments: preprint

Unravelling the fluorescence kinetics of light-harvesting proteins with simulated measurements

Authors: Callum Gray, Lekshmi Kailas, Peter G. Adams, Christopher D. P. Duffy

Abstract: The plant light-harvesting pigment-protein complex LHCII is the major antenna sub-unit of PSII and is generally (though not universally) accepted to play a role in photoprotective energy dissipation under high light conditions, a process known Non-Photochemical Quenching (NPQ). The underlying mechanisms of energy trapping and dissipation within LHCII are still debated. Various proposed models diff… ▽ More The plant light-harvesting pigment-protein complex LHCII is the major antenna sub-unit of PSII and is generally (though not universally) accepted to play a role in photoprotective energy dissipation under high light conditions, a process known Non-Photochemical Quenching (NPQ). The underlying mechanisms of energy trapping and dissipation within LHCII are still debated. Various proposed models differ considerably in their molecular and kinetic detail, but are often based on different interpretations of very similar transient absorption measurements of isolated complexes. Here we present a simulated measurement of the fluorescence decay kinetics of quenched LHCII aggregates to determine whether this relatively simple measurement can discriminate between different potential NPQ mechanisms. We simulate not just the underlying physics (excitation, energy migration, quenching and singlet-singlet annihilation) but also the signal detection and typical experimental data analysis. Comparing this to a selection of published fluorescence decay kinetics we find that: (1) Different proposed quenching mechanisms produce noticeably different fluorescence kinetics even at low (annihilation free) excitation density, though the degree of difference is dependent on pulse width. (2) Measured decay kinetics are consistent with most LHCII trimers becoming relatively slow excitation quenchers. A small sub-population of very fast quenchers produces kinetics which do not resemble any observed measurement. (3) It is necessary to consider at least two distinct quenching mechanisms in order to accurately reproduce experimental kinetics, supporting the idea that NPQ is not a simple binary switch switch. △ Less

Submitted 26 July, 2023; originally announced July 2023.

Comments: 5 figures, original research article submitted to BBA-Bioenergetics

Generating EDU Extracts for Plan-Guided Summary Re-Ranking

Authors: Griffin Adams, Alexander R. Fabbri, Faisal Ladhak, Kathleen McKeown, Noémie Elhadad

Abstract: Two-step approaches, in which summary candidates are generated-then-reranked to return a single summary, can improve ROUGE scores over the standard single-step approach. Yet, standard decoding methods (i.e., beam search, nucleus sampling, and diverse beam search) produce candidates with redundant, and often low quality, content. In this paper, we design a novel method to generate candidates for re… ▽ More Two-step approaches, in which summary candidates are generated-then-reranked to return a single summary, can improve ROUGE scores over the standard single-step approach. Yet, standard decoding methods (i.e., beam search, nucleus sampling, and diverse beam search) produce candidates with redundant, and often low quality, content. In this paper, we design a novel method to generate candidates for re-ranking that addresses these issues. We ground each candidate abstract on its own unique content plan and generate distinct plan-guided abstracts using a model's top beam. More concretely, a standard language model (a BART LM) auto-regressively generates elemental discourse unit (EDU) content plans with an extractive copy mechanism. The top K beams from the content plan generator are then used to guide a separate LM, which produces a single abstractive candidate for each distinct plan. We apply an existing re-ranker (BRIO) to abstractive candidates generated from our method, as well as baseline decoding methods. We show large relevance improvements over previously published methods on widely used single document news article corpora, with ROUGE-2 F1 gains of 0.88, 2.01, and 0.38 on CNN / Dailymail, NYT, and Xsum, respectively. A human evaluation on CNN / DM validates these results. Similarly, on 1k samples from CNN / DM, we show that prompting GPT-3 to follow EDU plans outperforms sampling-based methods by 1.05 ROUGE-2 F1 points. Code to generate and realize plans is available at https://github.com/griff4692/edu-sum. △ Less

Submitted 28 May, 2023; originally announced May 2023.

Comments: ACL 2023

What are the Desired Characteristics of Calibration Sets? Identifying Correlates on Long Form Scientific Summarization

Authors: Griffin Adams, Bichlien H Nguyen, Jake Smith, Yingce Xia, Shufang Xie, Anna Ostropolets, Budhaditya Deb, Yuan-Jyue Chen, Tristan Naumann, Noémie Elhadad

Abstract: Summarization models often generate text that is poorly calibrated to quality metrics because they are trained to maximize the likelihood of a single reference (MLE). To address this, recent work has added a calibration step, which exposes a model to its own ranked outputs to improve relevance or, in a separate line of work, contrasts positive and negative sets to improve faithfulness. While effec… ▽ More Summarization models often generate text that is poorly calibrated to quality metrics because they are trained to maximize the likelihood of a single reference (MLE). To address this, recent work has added a calibration step, which exposes a model to its own ranked outputs to improve relevance or, in a separate line of work, contrasts positive and negative sets to improve faithfulness. While effective, much of this work has focused on how to generate and optimize these sets. Less is known about why one setup is more effective than another. In this work, we uncover the underlying characteristics of effective sets. For each training instance, we form a large, diverse pool of candidates and systematically vary the subsets used for calibration fine-tuning. Each selection strategy targets distinct aspects of the sets, such as lexical diversity or the size of the gap between positive and negatives. On three diverse scientific long-form summarization datasets (spanning biomedical, clinical, and chemical domains), we find, among others, that faithfulness calibration is optimal when the negative sets are extractive and more likely to be generated, whereas for relevance calibration, the metric margin between candidates should be maximized and surprise--the disagreement between model and metric defined candidate rankings--minimized. Code to create, select, and optimize calibration sets is available at https://github.com/griff4692/calibrating-summaries △ Less

Submitted 12 May, 2023; originally announced May 2023.

Comments: ACL 2023

Marginal deformations of Calabi-Yau hypersurface hybrids with (2,2) supersymmetry

Authors: Griffen Adams, Ilarion V. Melnikov

Abstract: We study two-dimensional non-linear sigma models with (2,2) supersymmetry and a holomorphic superpotential that are believed to flow to unitary compact (2,2) superconformal theories with equal left and right central charges c=9. The SCFTs have a set of marginal deformations, and some of these can be realized as deformations of parameters of the UV theory, making it possible to apply techniques suc… ▽ More We study two-dimensional non-linear sigma models with (2,2) supersymmetry and a holomorphic superpotential that are believed to flow to unitary compact (2,2) superconformal theories with equal left and right central charges c=9. The SCFTs have a set of marginal deformations, and some of these can be realized as deformations of parameters of the UV theory, making it possible to apply techniques such as localization to probe the deformations of the SCFT in terms of a UV Lagrangian. In this work we describe the UV lifts of the remaining SCFT infinitesimal deformations, the so-called non-toric and non-polynomial deformations. Our UV theories naturally arise as geometric phases of gauged linear sigma models, and it may be possible to extend our results to find lifts of all SCFT deformations to the gauged linear sigma model. △ Less

Submitted 10 May, 2023; originally announced May 2023.

Comments: 34 pages, tikz figures

A Meta-Evaluation of Faithfulness Metrics for Long-Form Hospital-Course Summarization

Authors: Griffin Adams, Jason Zucker, Noémie Elhadad

Abstract: Long-form clinical summarization of hospital admissions has real-world significance because of its potential to help both clinicians and patients. The faithfulness of summaries is critical to their safe usage in clinical settings. To better understand the limitations of abstractive systems, as well as the suitability of existing evaluation metrics, we benchmark faithfulness metrics against fine-gr… ▽ More Long-form clinical summarization of hospital admissions has real-world significance because of its potential to help both clinicians and patients. The faithfulness of summaries is critical to their safe usage in clinical settings. To better understand the limitations of abstractive systems, as well as the suitability of existing evaluation metrics, we benchmark faithfulness metrics against fine-grained human annotations for model-generated summaries of a patient's Brief Hospital Course. We create a corpus of patient hospital admissions and summaries for a cohort of HIV patients, each with complex medical histories. Annotators are presented with summaries and source notes, and asked to categorize manually highlighted summary elements (clinical entities like conditions and medications as well as actions like "following up") into one of three categories: ``Incorrect,'' ``Missing,'' and ``Not in Notes.'' We meta-evaluate a broad set of proposed faithfulness metrics and, across metrics, explore the importance of domain adaptation (e.g. the impact of in-domain pre-training and metric fine-tuning), the use of source-summary alignments, and the effects of distilling a single metric from an ensemble of pre-existing metrics. Off-the-shelf metrics with no exposure to clinical text correlate well yet overly rely on summary extractiveness. As a practical guide to long-form clinical narrative summarization, we find that most metrics correlate best to human judgments when provided with one summary sentence at a time and a minimal set of relevant source context. △ Less

Submitted 7 March, 2023; originally announced March 2023.

Comments: Preprint

Worst Case Resistance Testing: A Nonresponse Bias Solution for Today's Behavioral Research Realities

Authors: Stephen L. France, Frank G. Adams, V. Myles Landers

Abstract: This study proposes a method of nonresponse assessment based on meta-analytical file-drawer techniques, also known as worst-case resistance testing (WCRT), and suitable for a wide range of data collection scenarios. A general method is devised to estimate the number of significantly different nonrespondents it would take to significantly alter the results of an analysis. Estimates of nonrespondent… ▽ More This study proposes a method of nonresponse assessment based on meta-analytical file-drawer techniques, also known as worst-case resistance testing (WCRT), and suitable for a wide range of data collection scenarios. A general method is devised to estimate the number of significantly different nonrespondents it would take to significantly alter the results of an analysis. Estimates of nonrespondents can be plotted against effect sizes using "n-curves", with similar interpretation to p-curves or power curves. Variants of the general method are derived for tests of means and correlations. A sample using a well-established survey instrument from previous behavioral research is used to test the method. The results suggest that employing worst-case resistance testing can be used on its own or in conjunction with wave analysis to precisely flag nonresponse risks. △ Less

Submitted 19 January, 2023; originally announced January 2023.

MSC Class: 62P15 ACM Class: G.3; J.4

Learning to Revise References for Faithful Summarization

Authors: Griffin Adams, Han-Chin Shing, Qing Sun, Christopher Winestock, Kathleen McKeown, Noémie Elhadad

Abstract: In real-world scenarios with naturally occurring datasets, reference summaries are noisy and may contain information that cannot be inferred from the source text. On large news corpora, removing low quality samples has been shown to reduce model hallucinations. Yet, for smaller, and/or noisier corpora, filtering is detrimental to performance. To improve reference quality while retaining all data,… ▽ More In real-world scenarios with naturally occurring datasets, reference summaries are noisy and may contain information that cannot be inferred from the source text. On large news corpora, removing low quality samples has been shown to reduce model hallucinations. Yet, for smaller, and/or noisier corpora, filtering is detrimental to performance. To improve reference quality while retaining all data, we propose a new approach: to selectively re-write unsupported reference sentences to better reflect source data. We automatically generate a synthetic dataset of positive and negative revisions by corrupting supported sentences and learn to revise reference sentences with contrastive learning. The intensity of revisions is treated as a controllable attribute so that, at inference, diverse candidates can be over-generated-then-rescored to balance faithfulness and abstraction. To test our methods, we extract noisy references from publicly available MIMIC-III discharge summaries for the task of hospital-course summarization, and vary the data on which models are trained. According to metrics and human evaluation, models trained on revised clinical references are much more faithful, informative, and fluent than models trained on original or filtered data. △ Less

Submitted 11 October, 2022; v1 submitted 13 April, 2022; originally announced April 2022.

Comments: Findings of EMNLP 2022

The 1.28 GHz MeerKAT Galactic Center Mosaic

Authors: I. Heywood, I. Rammala, F. Camilo, W. D. Cotton, F. Yusef-Zadeh, T. D. Abbott, R. M. Adam, G. Adams, M. A. Aldera, K. M. B. Asad, E. F. Bauermeister, T. G. H. Bennett, H. L. Bester, W. A. Bode, D. H. Botha, A. G. Botha, L. R. S. Brederode, S. Buchner, J. P. Burger, T. Cheetham, D. I. L. de Villiers, M. A. Dikgale-Mahlakoana, L. J. du Toit, S. W. P. Esterhuyse, B. L. Fanaroff , et al. (86 additional authors not shown)

Abstract: The inner $\sim$200 pc region of the Galaxy contains a 4 million M$_{\odot}$ supermassive black hole (SMBH), significant quantities of molecular gas, and star formation and cosmic ray energy densities that are roughly two orders of magnitude higher than the corresponding levels in the Galactic disk. At a distance of only 8.2 kpc, the region presents astronomers with a unique opportunity to study a… ▽ More The inner $\sim$200 pc region of the Galaxy contains a 4 million M$_{\odot}$ supermassive black hole (SMBH), significant quantities of molecular gas, and star formation and cosmic ray energy densities that are roughly two orders of magnitude higher than the corresponding levels in the Galactic disk. At a distance of only 8.2 kpc, the region presents astronomers with a unique opportunity to study a diverse range of energetic astrophysical phenomena, from stellar objects in extreme environments, to the SMBH and star-formation driven feedback processes that are known to influence the evolution of galaxies as a whole. We present a new survey of the Galactic center conducted with the South African MeerKAT radio telescope. Radio imaging offers a view that is unaffected by the large quantities of dust that obscure the region at other wavelengths, and a scene of striking complexity is revealed. We produce total intensity and spectral index mosaics of the region from 20 pointings (144 hours on-target in total), covering 6.5 square degrees with an angular resolution of 4$"$,at a central frequency of 1.28 GHz. Many new features are revealed for the first time due to a combination of MeerKAT's high sensitivity, exceptional $u,v$-plane coverage, and geographical vantage point. We highlight some initial survey results, including new supernova remnant candidates, many new non-thermal filament complexes, and enhanced views of the Radio Arc Bubble, Sgr A and Sgr B regions. This project is a SARAO public legacy survey, and the image products are made available with this article. △ Less

Submitted 27 January, 2022; v1 submitted 25 January, 2022; originally announced January 2022.

Comments: 29 pages, 15 figures, 3 tables, accepted for publication in the The Astrophysical Journal. Replacement arXiv version with higher quality figures

The MeerKAT Galaxy Cluster Legacy Survey I. Survey Overview and Highlights

Authors: K. Knowles, W. D. Cotton, L. Rudnick, F. Camilo, S. Goedhart, R. Deane, M. Ramatsoku, M. F. Bietenholz, M. Brüggen, C. Button, H. Chen, J. O. Chibueze, T. E. Clarke, F. de Gasperin, R. Ianjamasimanana, G. I. G. Józsa, M. Hilton, K. C. Kesebonye, K. Kolokythas, R. C. Kraan-Korteweg, G. Lawrie, M. Lochner, S. I. Loubser, P. Marchegiani, N. Mhlahlo , et al. (126 additional authors not shown)

Abstract: MeerKAT's large number of antennas, spanning 8 km with a densely packed 1 km core, create a powerful instrument for wide-area surveys, with high sensitivity over a wide range of angular scales. The MeerKAT Galaxy Cluster Legacy Survey (MGCLS) is a programme of long-track MeerKAT L-band (900-1670 MHz) observations of 115 galaxy clusters, observed for $\sim$6-10 hours each in full polarisation. The… ▽ More MeerKAT's large number of antennas, spanning 8 km with a densely packed 1 km core, create a powerful instrument for wide-area surveys, with high sensitivity over a wide range of angular scales. The MeerKAT Galaxy Cluster Legacy Survey (MGCLS) is a programme of long-track MeerKAT L-band (900-1670 MHz) observations of 115 galaxy clusters, observed for $\sim$6-10 hours each in full polarisation. The first legacy product data release (DR1), made available with this paper, includes the MeerKAT visibilities, basic image cubes at $\sim$8" resolution, and enhanced spectral and polarisation image cubes at $\sim$8" and 15" resolutions. Typical sensitivities for the full-resolution MGCLS image products are $\sim$3-5 μJy/beam. The basic cubes are full-field and span 4 deg^2. The enhanced products consist of the inner 1.44 deg^2 field of view, corrected for the primary beam. The survey is fully sensitive to structures up to $\sim$10' scales and the wide bandwidth allows spectral and Faraday rotation mapping. HI mapping at 209 kHz resolution can be done at $0<z<0.09$ and $0.19<z<0.48$. In this paper, we provide an overview of the survey and DR1 products, including caveats for usage. We present some initial results from the survey, both for their intrinsic scientific value and to highlight the capabilities for further exploration with these data. These include a primary beam-corrected compact source catalogue of $\sim$626,000 sources for the full survey, and an optical/infrared cross-matched catalogue for compact sources in Abell 209 and Abell S295. We examine dust unbiased star-formation rates as a function of clustercentric radius in Abell 209 and present a catalogue of 99 diffuse cluster sources (56 are new), some of which have no suitable characterisation. We also highlight some of the radio galaxies which challenge current paradigms and present first results from HI studies of four targets. △ Less

Submitted 10 November, 2021; originally announced November 2021.

Comments: 43 pages, 33 figures, 6 tables. Accepted for publication in Astronomy and Astrophysics

Investigating Crowdsourcing Protocols for Evaluating the Factual Consistency of Summaries

Authors: Xiangru Tang, Alexander Fabbri, Haoran Li, Ziming Mao, Griffin Thomas Adams, Borui Wang, Asli Celikyilmaz, Yashar Mehdad, Dragomir Radev

Abstract: Current pre-trained models applied to summarization are prone to factual inconsistencies which either misrepresent the source text or introduce extraneous information. Thus, comparing the factual consistency of summaries is necessary as we develop improved models. However, the optimal human evaluation setup for factual consistency has not been standardized. To address this issue, we crowdsourced e… ▽ More Current pre-trained models applied to summarization are prone to factual inconsistencies which either misrepresent the source text or introduce extraneous information. Thus, comparing the factual consistency of summaries is necessary as we develop improved models. However, the optimal human evaluation setup for factual consistency has not been standardized. To address this issue, we crowdsourced evaluations for factual consistency using the rating-based Likert scale and ranking-based Best-Worst Scaling protocols, on 100 articles from each of the CNN-Daily Mail and XSum datasets over four state-of-the-art models, to determine the most reliable evaluation framework. We find that ranking-based protocols offer a more reliable measure of summary quality across datasets, while the reliability of Likert ratings depends on the target dataset and the evaluation design. Our crowdsourcing templates and summary evaluations will be publicly available to facilitate future research on factual consistency in summarization. △ Less

Submitted 9 July, 2022; v1 submitted 19 September, 2021; originally announced September 2021.

What's in a Summary? Laying the Groundwork for Advances in Hospital-Course Summarization

Authors: Griffin Adams, Emily Alsentzer, Mert Ketenci, Jason Zucker, Noémie Elhadad

Abstract: Summarization of clinical narratives is a long-standing research problem. Here, we introduce the task of hospital-course summarization. Given the documentation authored throughout a patient's hospitalization, generate a paragraph that tells the story of the patient admission. We construct an English, text-to-text dataset of 109,000 hospitalizations (2M source notes) and their corresponding summary… ▽ More Summarization of clinical narratives is a long-standing research problem. Here, we introduce the task of hospital-course summarization. Given the documentation authored throughout a patient's hospitalization, generate a paragraph that tells the story of the patient admission. We construct an English, text-to-text dataset of 109,000 hospitalizations (2M source notes) and their corresponding summary proxy: the clinician-authored "Brief Hospital Course" paragraph written as part of a discharge note. Exploratory analyses reveal that the BHC paragraphs are highly abstractive with some long extracted fragments; are concise yet comprehensive; differ in style and content organization from the source notes; exhibit minimal lexical cohesion; and represent silver-standard references. Our analysis identifies multiple implications for modeling this complex, multi-document summarization task. △ Less

Submitted 12 April, 2021; originally announced May 2021.

Comments: NAACL 2021

Resolving Implicit Coordination in Multi-Agent Deep Reinforcement Learning with Deep Q-Networks & Game Theory

Authors: Griffin Adams, Sarguna Janani Padmanabhan, Shivang Shekhar

Abstract: We address two major challenges of implicit coordination in multi-agent deep reinforcement learning: non-stationarity and exponential growth of state-action space, by combining Deep-Q Networks for policy learning with Nash equilibrium for action selection. Q-values proxy as payoffs in Nash settings, and mutual best responses define joint action selection. Coordination is implicit because multiple/… ▽ More We address two major challenges of implicit coordination in multi-agent deep reinforcement learning: non-stationarity and exponential growth of state-action space, by combining Deep-Q Networks for policy learning with Nash equilibrium for action selection. Q-values proxy as payoffs in Nash settings, and mutual best responses define joint action selection. Coordination is implicit because multiple/no Nash equilibria are resolved deterministically. We demonstrate that knowledge of game type leads to an assumption of mirrored best responses and faster convergence than Nash-Q. Specifically, the Friend-or-Foe algorithm demonstrates signs of convergence to a Set Controller which jointly chooses actions for two agents. This encouraging given the highly unstable nature of decentralized coordination over joint actions. Inspired by the dueling network architecture, which decouples the Q-function into state and advantage streams, as well as residual networks, we learn both a single and joint agent representation, and merge them via element-wise addition. This simplifies coordination by recasting it is as learning a residual function. We also draw high level comparative insights on key MADRL and game theoretic variables: competitive vs. cooperative, asynchronous vs. parallel learning, greedy versus socially optimal Nash equilibria tie breaking, and strategies for the no Nash equilibrium case. We evaluate on 3 custom environments written in Python using OpenAI Gym: a Predator Prey environment, an alternating Warehouse environment, and a Synchronization environment. Each environment requires successively more coordination to achieve positive rewards. △ Less

Submitted 8 December, 2020; originally announced December 2020.

Comments: 8 pages

Zero-Shot Clinical Acronym Expansion via Latent Meaning Cells

Authors: Griffin Adams, Mert Ketenci, Shreyas Bhave, Adler Perotte, Noémie Elhadad

Abstract: We introduce Latent Meaning Cells, a deep latent variable model which learns contextualized representations of words by combining local lexical context and metadata. Metadata can refer to granular context, such as section type, or to more global context, such as unique document ids. Reliance on metadata for contextualized representation learning is apropos in the clinical domain where text is semi… ▽ More We introduce Latent Meaning Cells, a deep latent variable model which learns contextualized representations of words by combining local lexical context and metadata. Metadata can refer to granular context, such as section type, or to more global context, such as unique document ids. Reliance on metadata for contextualized representation learning is apropos in the clinical domain where text is semi-structured and expresses high variation in topics. We evaluate the LMC model on the task of zero-shot clinical acronym expansion across three datasets. The LMC significantly outperforms a diverse set of baselines at a fraction of the pre-training cost and learns clinically coherent representations. We demonstrate that not only is metadata itself very helpful for the task, but that the LMC inference algorithm provides an additional large benefit. △ Less

Submitted 12 November, 2020; v1 submitted 28 September, 2020; originally announced October 2020.

Comments: To appear in Proceedings Track for Machine Learning for Health (ML4H) Workshop at NeurIPS (2020)

The MeerKAT Telescope as a Pulsar Facility: System verification and early science results from MeerTime

Authors: M. Bailes, A. Jameson, F. Abbate, E. D. Barr, N. D. R. Bhat, L. Bondonneau, M. Burgay, S. J. Buchner, F. Camilo, D. J. Champion, I. Cognard, P. B. Demorest, P. C. C. Freire, T. Gautam, M. Geyer, J. M. Griessmeier, L. Guillemot, H. Hu, F. Jankowski, S. Johnston, A. Karastergiou, R. Karuppusamy, D. Kaur, M. J. Keith, M. Kramer , et al. (50 additional authors not shown)

Abstract: We describe system verification tests and early science results from the pulsar processor (PTUSE) developed for the newly-commissioned 64-dish SARAO MeerKAT radio telescope in South Africa. MeerKAT is a high-gain (~2.8 K/Jy) low-system temperature (~18 K at 20cm) radio array that currently operates from 580-1670 MHz and can produce tied-array beams suitable for pulsar observations. This paper pres… ▽ More We describe system verification tests and early science results from the pulsar processor (PTUSE) developed for the newly-commissioned 64-dish SARAO MeerKAT radio telescope in South Africa. MeerKAT is a high-gain (~2.8 K/Jy) low-system temperature (~18 K at 20cm) radio array that currently operates from 580-1670 MHz and can produce tied-array beams suitable for pulsar observations. This paper presents results from the MeerTime Large Survey Project and commissioning tests with PTUSE. Highlights include observations of the double pulsar J0737-3039A, pulse profiles from 34 millisecond pulsars from a single 2.5h observation of the Globular cluster Terzan 5, the rotation measure of Ter5O, a 420-sigma giant pulse from the Large Magellanic Cloud pulsar PSR J0540-6919, and nulling identified in the slow pulsar PSR J0633-2015. One of the key design specifications for MeerKAT was absolute timing errors of less than 5 ns using their novel precise time system. Our timing of two bright millisecond pulsars confirm that MeerKAT delivers exceptional timing. PSR J2241-5236 exhibits a jitter limit of <4 ns per hour whilst timing of PSR J1909-3744 over almost 11 months yields an rms residual of 66 ns with only 4 min integrations. Our results confirm that the MeerKAT is an exceptional pulsar telescope. The array can be split into four separate sub-arrays to time over 1000 pulsars per day and the future deployment of S-band (1750-3500 MHz) receivers will further enhance its capabilities. △ Less

Submitted 28 May, 2020; originally announced May 2020.

Comments: 20 pages, 16 Figures, 4 Tables, accepted for publication in PASA

A Functional Decomposition of Finite Bandwidth Reproducing Kernel Hilbert Spaces

Authors: Gregory T. Adams, Nathan A. Wagner

Abstract: In this work, we consider "finite bandwidth" reproducing kernel Hilbert spaces which have orthonormal bases of the form $f_n(z)=z^n \prod_{j=1}^J \left( 1 - a_{n}w_j z \right)$, where $w_1 ,w_2, \ldots w_J $ are distinct points on the circle $\mathbb{T}$ and $\{ a_n \}$ is a sequence of complex numbers with limit $1$. We provide general conditions based on a matrix recursion that guarantee such sp… ▽ More In this work, we consider "finite bandwidth" reproducing kernel Hilbert spaces which have orthonormal bases of the form $f_n(z)=z^n \prod_{j=1}^J \left( 1 - a_{n}w_j z \right)$, where $w_1 ,w_2, \ldots w_J $ are distinct points on the circle $\mathbb{T}$ and $\{ a_n \}$ is a sequence of complex numbers with limit $1$. We provide general conditions based on a matrix recursion that guarantee such spaces contain a functional multiple of the Hardy space. Then we apply this general method to obtain strong results for finite bandwidth spaces when $\lim_{n\rightarrow \infty} n (1-a_n)=p$. In particular, we show that point evaluation can be extended boundedly to precisely $J$ additional points on $\mathbb{T}$ and we obtain an explicit functional decomposition of these spaces for $p>1/2$ in analogy with a previous result in the tridiagonal case due to Adams and McGuire. We also prove that multiplication by $z$ is a bounded operator on these spaces and that they contain the polynomials. △ Less

Submitted 28 August, 2019; originally announced August 2019.

Comments: 19 pages with references

MSC Class: 47

Journal ref: Oper. Matrices 15 (2021), no. 4, 1521-1539

TIFTI: A Framework for Extracting Drug Intervals from Longitudinal Clinic Notes

Authors: Monica Agrawal, Griffin Adams, Nathan Nussbaum, Benjamin Birnbaum

Abstract: Oral drugs are becoming increasingly common in oncology care. In contrast to intravenous chemotherapy, which is administered in the clinic and carefully tracked via structure electronic health records (EHRs), oral drug treatment is self-administered and therefore not tracked as well. Often, the details of oral cancer treatment occur only in unstructured clinic notes. Extracting this information is… ▽ More Oral drugs are becoming increasingly common in oncology care. In contrast to intravenous chemotherapy, which is administered in the clinic and carefully tracked via structure electronic health records (EHRs), oral drug treatment is self-administered and therefore not tracked as well. Often, the details of oral cancer treatment occur only in unstructured clinic notes. Extracting this information is critical to understanding a patient's treatment history. Yet, this a challenging task because treatment intervals must be inferred longitudinally from both explicit mentions in the text as well as from document timestamps. In this work, we present TIFTI (Temporally Integrated Framework for Treatment Intervals), a robust framework for extracting oral drug treatment intervals from a patient's unstructured notes. TIFTI leverages distinct sources of temporal information by breaking the problem down into two separate subtasks: document-level sequence labeling and date extraction. On a labeled dataset of metastatic renal-cell carcinoma (RCC) patients, it exactly matched the labeled start date in 46% of the examples (86% of the examples within 30 days), and it exactly matched the labeled end date in 52% of the examples (78% of the examples within 30 days). Without retraining, the model achieved a similar level of performance on a labeled dataset of advanced non-small-cell lung cancer (NSCLC) patients. △ Less

Submitted 3 December, 2018; v1 submitted 30 November, 2018; originally announced November 2018.

Comments: Machine Learning for Health (ML4H) Workshop at NeurIPS 2018 arXiv:1811.07216

Report number: ML4H/2018/69

Revival of the magnetar PSR J1622-4950: observations with MeerKAT, Parkes, XMM-Newton, Swift, Chandra, and NuSTAR

Authors: F. Camilo, P. Scholz, M. Serylak, S. Buchner, M. Merryfield, V. M. Kaspi, R. F. Archibald, M. Bailes, A. Jameson, W. van Straten, J. Sarkissian, J. E. Reynolds, S. Johnston, G. Hobbs, T. D. Abbott, R. M. Adam, G. B. Adams, T. Alberts, R. Andreas, K. M. B. Asad, D. E. Baker, T. Baloyi, E. F. Bauermeister, T. Baxana, T. G. H. Bennett , et al. (183 additional authors not shown)

Abstract: New radio (MeerKAT and Parkes) and X-ray (XMM-Newton, Swift, Chandra, and NuSTAR) observations of PSR J1622-4950 indicate that the magnetar, in a quiescent state since at least early 2015, reactivated between 2017 March 19 and April 5. The radio flux density, while variable, is approximately 100x larger than during its dormant state. The X-ray flux one month after reactivation was at least 800x la… ▽ More New radio (MeerKAT and Parkes) and X-ray (XMM-Newton, Swift, Chandra, and NuSTAR) observations of PSR J1622-4950 indicate that the magnetar, in a quiescent state since at least early 2015, reactivated between 2017 March 19 and April 5. The radio flux density, while variable, is approximately 100x larger than during its dormant state. The X-ray flux one month after reactivation was at least 800x larger than during quiescence, and has been decaying exponentially on a 111+/-19 day timescale. This high-flux state, together with a radio-derived rotational ephemeris, enabled for the first time the detection of X-ray pulsations for this magnetar. At 5%, the 0.3-6 keV pulsed fraction is comparable to the smallest observed for magnetars. The overall pulsar geometry inferred from polarized radio emission appears to be broadly consistent with that determined 6-8 years earlier. However, rotating vector model fits suggest that we are now seeing radio emission from a different location in the magnetosphere than previously. This indicates a novel way in which radio emission from magnetars can differ from that of ordinary pulsars. The torque on the neutron star is varying rapidly and unsteadily, as is common for magnetars following outburst, having changed by a factor of 7 within six months of reactivation. △ Less

Submitted 5 April, 2018; originally announced April 2018.

Comments: Published in ApJ (2018 April 5); 13 pages, 4 figures

Journal ref: ApJ 856 (2018) 180

Predicting Chronic Disease Hospitalizations from Electronic Health Records: An Interpretable Classification Approach

Authors: Theodora S. Brisimi, Tingting Xu, Taiyao Wang, Wuyang Dai, William G. Adams, Ioannis Ch. Paschalidis

Abstract: Urban living in modern large cities has significant adverse effects on health, increasing the risk of several chronic diseases. We focus on the two leading clusters of chronic disease, heart disease and diabetes, and develop data-driven methods to predict hospitalizations due to these conditions. We base these predictions on the patients' medical history, recent and more distant, as described in t… ▽ More Urban living in modern large cities has significant adverse effects on health, increasing the risk of several chronic diseases. We focus on the two leading clusters of chronic disease, heart disease and diabetes, and develop data-driven methods to predict hospitalizations due to these conditions. We base these predictions on the patients' medical history, recent and more distant, as described in their Electronic Health Records (EHR). We formulate the prediction problem as a binary classification problem and consider a variety of machine learning methods, including kernelized and sparse Support Vector Machines (SVM), sparse logistic regression, and random forests. To strike a balance between accuracy and interpretability of the prediction, which is important in a medical setting, we propose two novel methods: K-LRT, a likelihood ratio test-based method, and a Joint Clustering and Classification (JCC) method which identifies hidden patient clusters and adapts classifiers to each cluster. We develop theoretical out-of-sample guarantees for the latter method. We validate our algorithms on large datasets from the Boston Medical Center, the largest safety-net hospital system in New England. △ Less

Submitted 3 January, 2018; originally announced January 2018.

MSC Class: (Primary) 62H30; 65Kxx (Secondary) ACM Class: I.5.2; I.5.3; J.3

Measurements of the Separated Longitudinal Structure Function F_L from Hydrogen and Deuterium Targets at Low Q^2

Authors: V. Tvaskis, A. Tvaskis, I. Niculescu, D. Abbott, G. S. Adams, A. Afanasev, A. Ahmidouch, T. Angelescu, J. Arrington, R. Asaturyan, S. Avery, O. K. Baker, N. Benmouna, B. L. Berman, A. Biselli, H. P. Blok, W. U. Boeglin, P. E. Bosted, E. Brash, H. Breuer, G. Chang, N. Chant, M. E. Christy, S. H. Connell, M. M. Dalton , et al. (78 additional authors not shown)

Abstract: Structure functions, as measured in lepton-nucleon scattering, have proven to be very useful in studying the quark dynamics within the nucleon. However, it is experimentally difficult to separately determine the longitudinal and transverse structure functions, and consequently there are substantially less data available for the longitudinal structure function in particular. Here we present separat… ▽ More Structure functions, as measured in lepton-nucleon scattering, have proven to be very useful in studying the quark dynamics within the nucleon. However, it is experimentally difficult to separately determine the longitudinal and transverse structure functions, and consequently there are substantially less data available for the longitudinal structure function in particular. Here we present separated structure functions for hydrogen and deuterium at low four--momentum transfer squared, Q^2< 1 GeV^2, and compare these with parton distribution parameterizations and a k_T factorization approach. While differences are found, the parameterizations generally agree with the data even at the very low Q^2 scale of the data. The deuterium data show a smaller longitudinal structure function, and smaller ratio of longitudinal to transverse cross section R, than the proton. This suggests either an unexpected difference in R for the proton and neutron or a suppression of the gluonic distribution in nuclei. △ Less

Submitted 8 June, 2016; originally announced June 2016.

Comments: 11 pages, 7 figures, 2 data tables

Journal ref: Phys. Rev. C 97, 045204 (2018)

Estimating demographic parameters using a combination of known-fate and open N-mixture models

Authors: Joshua H. Schmidt, Devin S. Johnson, Mark S. Lindberg, Layne G. Adams

Abstract: 1. Accurate estimates of demographic parameters are required to infer appropriate ecological relationships and inform management actions. Recently developed N-mixture models use count data from unmarked individuals to estimate demographic parameters, but a joint approach combining the strengths of both analytical tools has not been developed. 2. We present an integrated model combining known-fate… ▽ More 1. Accurate estimates of demographic parameters are required to infer appropriate ecological relationships and inform management actions. Recently developed N-mixture models use count data from unmarked individuals to estimate demographic parameters, but a joint approach combining the strengths of both analytical tools has not been developed. 2. We present an integrated model combining known-fate and open N-mixture models, allowing the estimation of detection probability, recruitment, and the joint estimation of survival. We first use a simulation study to evaluate the performance of the model relative to known values. We then provide an applied example using 4 years of wolf survival data consisting of relocations of radio-collared wolves within packs and counts of associated pack-mates. The model is implemented in both maximum-likelihood and Bayesian frameworks using a new R package kfdnm and the BUGS language. 3. The simulation results indicated that the integrated model was able to reliably recover parameters with no evidence of bias, and estimates were more precise under the joint model as expected. Results from the applied example indicated that the marked sample of wolves was biased towards individuals with higher apparent survival rates (including losses due to mortality and emigration) than the unmarked pack-mates, suggesting estimates of apparent survival based on joint estimation could be more representative of the overall population. Estimates of recruitment were similar to direct observations of pup production, and overlap of the credible intervals suggested no clear differences in recruitment rates. 4. Our integrated model is a practical approach for increasing the amount of information gained from future and existing radio-telemetry and other similar mark-resight datasets. △ Less

Submitted 10 February, 2015; originally announced February 2015.

Comments: 22 pages, 2 figures, 2 appendices

Journal ref: Ecology 96:2583-2589 (2015)

Updated measurements of absolute $D^+$ and $D^0$ hadronic branching fractions and $σ(e^+e^-\to D\overline{D})$ at $E_\mathrm{cm} = 3774$ MeV

Authors: CLEO Collaboration, G. Bonvicini, D. Cinabro M. J. Smith, P. Zhou, P. Naik, J. Rademacker, K. W. Edwards, R. A. Briere, H. Vogel, J. L. Rosner, J. P. Alexander, D. G. Cassel, R. Ehrlich, L. Gibbons, S. W. Gray, D. L. Hartill, B. K. Heltsley, D. L. Kreinick, V. E. Kuznetsov, J. R. Patterson, D. Peterson, D. Riley, A. Ryd, A. J. Sadoff, X. Shi , et al. (44 additional authors not shown)

Abstract: Utilizing the full CLEO-c data sample of 818 pb$^{-1}$ of $e^+e^-$ data taken at the $ψ(3770)$ resonance, we update our measurements of absolute hadronic branching fractions of charged and neutral $D$ mesons. We previously reportedresults from subsets of these data. Using a double tag technique we obtain branching fractions for three $D^0$ and six $D^+$ modes, including the reference branching fra… ▽ More Utilizing the full CLEO-c data sample of 818 pb$^{-1}$ of $e^+e^-$ data taken at the $ψ(3770)$ resonance, we update our measurements of absolute hadronic branching fractions of charged and neutral $D$ mesons. We previously reportedresults from subsets of these data. Using a double tag technique we obtain branching fractions for three $D^0$ and six $D^+$ modes, including the reference branching fractions $\mathcal{B} (D^0\to K^-π^+)=(3.934 \pm 0.021 \pm 0.061)\%$ and $\mathcal{B} (D^+ \to K^- π^+π^+)=(9.224 \pm 0.059 \pm 0.157)\%$. The uncertainties are statistical and systematic, respectively. In these measurements we include the effects of final-state radiation by allowing for additional unobserved photons in the final state, and the systematic errors include our estimates of the uncertainties of these effects. Furthermore, using an independent measurement of the luminosity, we obtain the cross sections $σ(e^+e^-\to D^0\overline{D}{}^0)=(3.607\pm 0.017 \pm 0.056) \ \mathrm{nb}$ and $σ(e^+e^-\to D^+D^-)=(2.882\pm 0.018 \pm 0.042) \ \mathrm{nb}$ at a center of mass energy, $E_\mathrm{cm} = 3774 \pm 1$ MeV. △ Less

Submitted 20 August, 2014; v1 submitted 24 December, 2013; originally announced December 2013.

Comments: 12 pages, 2 figures, and 6 tables. The Fig. 2 in this version corrects errors in plotting the fits in the original Fig. 2. Correcting these errors does not affect any result or text. We thank Dr. Maurice Garcia-Sciveres for pointing out this mistake

Report number: CLNS 13/2087, CLEO 13-02

Journal ref: Phys. Rev. D 89, 072002 (2014)

Improved Measurement of Absolute Hadronic Branching Fractions of the Ds+ Meson

Authors: CLEO Collaboration, P. U. E. Onyisi, G. Bonvicini, D. Cinabro, M. J. Smith, P. Zhou, P. Naik, J. Rademacker, K. W. Edwards, R. A. Briere, H. Vogel, J. L. Rosner, J. P. Alexander, D. G. Cassel, S. Das, R. Ehrlich, L. Gibbons, S. W. Gray, D. L. Hartill, B. K. Heltsley, D. L. Kreinick, V. E. Kuznetsov, J. R. Patterson, D. Peterson, D. Riley , et al. (45 additional authors not shown)

Abstract: The branching fractions of Ds meson decays serve to normalize many measurements of processes involving charm quarks. Using 586 pb^-1 of e+ e- collisions recorded at a center of mass energy of 4.17 GeV, we determine absolute branching fractions for 13 Ds decays in 16 reconstructed final states with a double tag technique. In particular we make a precise measurement of the branching fraction B(Ds ->… ▽ More The branching fractions of Ds meson decays serve to normalize many measurements of processes involving charm quarks. Using 586 pb^-1 of e+ e- collisions recorded at a center of mass energy of 4.17 GeV, we determine absolute branching fractions for 13 Ds decays in 16 reconstructed final states with a double tag technique. In particular we make a precise measurement of the branching fraction B(Ds -> K- K+ pi+) = (5.55 +- 0.14 +- 0.13)%, where the uncertainties are statistical and systematic respectively. We find a significantly reduced value of B(Ds -> pi+ pi0 eta') compared to the world average, and our results bring the inclusively and exclusively measured values of B(Ds -> eta' X)$ into agreement. We also search for CP-violating asymmetries in Ds decays and measure the cross-section of e+ e- -> Ds* Ds at Ecm = 4.17 GeV. △ Less

Submitted 14 September, 2013; v1 submitted 22 June, 2013; originally announced June 2013.

Comments: 12 pages, 3 figures. Submitted to Physical Review D. PDF also available through http://www.lns.cornell.edu/public/CLNS/2013/ v2: as published in PRD

Report number: CLNS 13/2086, CLEO 13-01

Journal ref: Phys. Rev. D 88, 032009 (2013)

Ending-based Strategies for Part-of-speech Tagging

Authors: Greg Adams, Beth Millar, Eric Neufeld, Tim Philip

Abstract: Probabilistic approaches to part-of-speech tagging rely primarily on whole-word statistics about word/tag combinations as well as contextual information. But experience shows about 4 per cent of tokens encountered in test sets are unknown even when the training set is as large as a million words. Unseen words are tagged using secondary strategies that exploit word features such as endings, capit… ▽ More Probabilistic approaches to part-of-speech tagging rely primarily on whole-word statistics about word/tag combinations as well as contextual information. But experience shows about 4 per cent of tokens encountered in test sets are unknown even when the training set is as large as a million words. Unseen words are tagged using secondary strategies that exploit word features such as endings, capitalizations and punctuation marks. In this work, word-ending statistics are primary and whole-word statistics are secondary. First, a tagger was trained and tested on word endings only. Subsequent experiments added back whole-word statistics for the words occurring most frequently in the training set. As grew larger, performance was expected to improve, in the limit performing the same as word-based taggers. Surprisingly, the ending-based tagger initially performed nearly as well as the word-based tagger; in the best case, its performance significantly exceeded that of the word-based tagger. Lastly, and unexpectedly, an effect of negative returns was observed - as grew larger, performance generally improved and then declined. By varying factors such as ending length and tag-list strategy, we achieved a success rate of 97.5 percent. △ Less

Submitted 27 February, 2013; originally announced February 2013.

Comments: Appears in Proceedings of the Tenth Conference on Uncertainty in Artificial Intelligence (UAI1994)

Report number: UAI-P-1994-PG-1-7

Updated Measurement of the Strong Phase in D0 --> K+pi- Decay Using Quantum Correlations in e+e- --> D0 D0bar at CLEO

Authors: CLEO Collaboration, D. M. Asner, G. Tatishvili, J. Y. Ge, D. H. Miller, I. P. J. Shipsey, B. Xin, G. S. Adams, J. Napolitano, K. M. Ecklund, Q. He, J. Insler, H. Muramatsu, L. J. Pearson, E. H. Thorndike, M. Artuso, S. Blusk, N. Horwitz, R. Mountain, T. Skwarnicki, S. Stone, J. C. Wang, L. M. Zhang, P. U. E. Onyisi, G. Bonvicini , et al. (50 additional authors not shown)

Abstract: We analyze a sample of 3 million quantum-correlated D0 D0bar pairs from 818 pb^-1 of e+e- collision data collected with the CLEO-c detector at E_cm = 3.77 GeV, to give an updated measurement of \cosδand a first determination of \sinδ, where δis the relative strong phase between doubly Cabibbo-suppressed D0 --> K+pi- and Cabibbo-favored D0bar --> K+pi- decay amplitudes. With no inputs from other ex… ▽ More We analyze a sample of 3 million quantum-correlated D0 D0bar pairs from 818 pb^-1 of e+e- collision data collected with the CLEO-c detector at E_cm = 3.77 GeV, to give an updated measurement of \cosδand a first determination of \sinδ, where δis the relative strong phase between doubly Cabibbo-suppressed D0 --> K+pi- and Cabibbo-favored D0bar --> K+pi- decay amplitudes. With no inputs from other experiments, we find \cosδ= 0.81 +0.22+0.07 -0.18-0.05, \sinδ= -0.01 +- 0.41 +- 0.04, and |δ| = 10 +28+13 -53-0 degrees. By including external measurements of mixing parameters, we find alternative values of \cosδ= 1.15 +0.19+0.00 -0.17-0.08, \sinδ= 0.56 +0.32+0.21 -0.31-0.20, and δ= (18 +11-17) degrees. Our results can be used to improve the world average uncertainty on the mixing parameter y by approximately 10%. △ Less

Submitted 7 November, 2012; v1 submitted 2 October, 2012; originally announced October 2012.

Comments: Minor revisions, version accepted by PRD

Report number: CLNS 12/2083; CLEO 12-01

Studies of the decays D^0 \rightarrow K_S^0K^-π^+ and D^0 \rightarrow K_S^0K^+π^-

Authors: CLEO Collaboration, J. Insler, H. Muramatsu, C. S. Park, L. J. Pearson, E. H. Thorndike, S. Ricciardi, C. Thomas, M. Artuso, S. Blusk, R. Mountain, T. Skwarnicki, S. Stone, J. C. Wang, L. M. Zhang, G. Bonvicini, D. Cinabro, M. J. Smith, P. Zhou, T. Gershon, P. Naik, J. Rademacker, K. W. Edwards, K. Randrianarivony, R. A. Briere , et al. (51 additional authors not shown)

Abstract: The first measurements of the coherence factor R_{K_S^0Kπ} and the average strong--phase difference δ^{K_S^0Kπ} in D^0 \to K_S^0 K^\mpπ^\pm decays are reported. These parameters can be used to improve the determination of the unitary triangle angle γ in B^- \rightarrow $\widetilde{D}K^-$ decays, where $\widetilde{D}$ is either a D^0 or a D^0-bar meson decaying to the same final state, and also in… ▽ More The first measurements of the coherence factor R_{K_S^0Kπ} and the average strong--phase difference δ^{K_S^0Kπ} in D^0 \to K_S^0 K^\mpπ^\pm decays are reported. These parameters can be used to improve the determination of the unitary triangle angle γ in B^- \rightarrow $\widetilde{D}K^-$ decays, where $\widetilde{D}$ is either a D^0 or a D^0-bar meson decaying to the same final state, and also in studies of charm mixing. The measurements of the coherence factor and strong-phase difference are made using quantum-correlated, fully-reconstructed D^0D^0-bar pairs produced in e^+e^- collisions at the ψ(3770) resonance. The measured values are R_{K_S^0Kπ} = 0.70 \pm 0.08 and δ^{K_S^0Kπ} = (0.1 \pm 15.7)$^\circ$ for an unrestricted kinematic region and R_{K*K} = 0.94 \pm 0.12 and δ^{K*K} = (-16.6 \pm 18.4)$^\circ$ for a region where the combined K_S^0 π^\pm invariant mass is within 100 MeV/c^2 of the K^{*}(892)^\pm mass. These results indicate a significant level of coherence in the decay. In addition, isobar models are presented for the two decays, which show the dominance of the K^*(892)^\pm resonance. The branching ratio {B}(D^0 \rightarrow K_S^0K^+π^-)/{B}(D^0 \rightarrow K_S^0K^-π^+) is determined to be 0.592 \pm 0.044 (stat.) \pm 0.018 (syst.), which is more precise than previous measurements. △ Less

Submitted 25 October, 2016; v1 submitted 16 March, 2012; originally announced March 2012.

Comments: 38 pages. Version 3 updated to include the erratum information. Errors corrected in Eqs (25), (26), 28). Fit results updated accordingly, and external inputs updated to latest best known values. Typo corrected in Eq(3)- no other consequences

Report number: CLNS 11/2081; CLEO 11-07

Journal ref: Phys. Rev. D 85 092016 2012

A KInetic Database for Astrochemistry (KIDA)

Authors: V. Wakelam, E. Herbst, J. -C. Loison, I. W. M. Smith, V. Chandrasekaran, B. Pavone, N. G. Adams, M. -C. Bacchus-Montabonel, A. Bergeat, K. Béroff, V. M. Bierbaum, M. Chabot, A. Dalgarno, E. F. van Dishoeck, A. Faure, W. D. Geppert, D. Gerlich, D. Galli, E. Hébrard, F. Hersant, K. M. Hickson, P. Honvault, S. J. Klippenstein, S. Le Picard, G. Nyman , et al. (9 additional authors not shown)

Abstract: We present a novel chemical database for gas-phase astrochemistry. Named the KInetic Database for Astrochemistry (KIDA), this database consists of gas-phase reactions with rate coefficients and uncertainties that will be vetted to the greatest extent possible. Submissions of measured and calculated rate coefficients are welcome, and will be studied by experts before inclusion into the database. Be… ▽ More We present a novel chemical database for gas-phase astrochemistry. Named the KInetic Database for Astrochemistry (KIDA), this database consists of gas-phase reactions with rate coefficients and uncertainties that will be vetted to the greatest extent possible. Submissions of measured and calculated rate coefficients are welcome, and will be studied by experts before inclusion into the database. Besides providing kinetic information for the interstellar medium, KIDA is planned to contain such data for planetary atmospheres and for circumstellar envelopes. Each year, a subset of the reactions in the database (kida.uva) will be provided as a network for the simulation of the chemistry of dense interstellar clouds with temperatures between 10 K and 300 K. We also provide a code, named Nahoon, to study the time-dependent gas-phase chemistry of 0D and 1D interstellar sources. △ Less

Submitted 27 January, 2012; originally announced January 2012.

Comments: Accepted for publication in ApJS

Amplitude analysis of D0->K+K-pi+pi-

Authors: M. Artuso, S. Blusk, R. Mountain, T. Skwarnicki, S. Stone, L. M. Zhang, T. Gershon, G. Bonvicini, D. Cinabro, A. Lincoln, M. J. Smith, P. Zhou, J. Zhu, P. Naik, J. Rademacker, D. M. Asner, K. W. Edwards, K. Randrianarivony, G. Tatishvili, R. A. Briere, H. Vogel, P. U. E. Onyisi, J. L. Rosner, J. P. Alexander, D. G. Cassel , et al. (51 additional authors not shown)

Abstract: The first flavor-tagged amplitude analysis of the decay D0 to the self-conjugate final state K+K-pi+pi- is presented. Data from the CLEO II.V, CLEO III, and CLEO-c detectors are used, from which around 3000 signal decays are selected. The three most significant amplitudes, which contribute to the model that best fits the data, are phirho0, K1(1270)+-K-+, and non-resonant K+K-pi+pi-. Separate ampli… ▽ More The first flavor-tagged amplitude analysis of the decay D0 to the self-conjugate final state K+K-pi+pi- is presented. Data from the CLEO II.V, CLEO III, and CLEO-c detectors are used, from which around 3000 signal decays are selected. The three most significant amplitudes, which contribute to the model that best fits the data, are phirho0, K1(1270)+-K-+, and non-resonant K+K-pi+pi-. Separate amplitude analyses of D0 and D0-bar candidates indicate no CP violation among the amplitudes at the level of 5% to 30% depending on the mode. In addition, the sensitivity to the CP-violating parameter gamma/phi3 of a sample of 2000 B+ -> D0-tilde(K+K-pi+pi-)K+ decays, where D0-tilde is a D0 or D0-bar, collected at LHCb or a future flavor facility, is estimated to be (11.3 +/- 0.3) degrees using the favored model. △ Less

Submitted 29 June, 2012; v1 submitted 27 January, 2012; originally announced January 2012.

Comments: 40 pages, 7 figures and submitted to Phys. Rev D

Report number: CLNS 11/2082; CLEO 11-08

Journal ref: Phys. Rev. D 85, 122002 (2012)

First Measurement of the Form Factors in the Decays D0 to rho- e+ nu_e and D+ to rho0 e+ nu_e

Authors: CLEO Collaboration, S. Dobbs, Z. Metreveli, K. K. Seth, A. Tomaradze, T. Xiao, L. Martin, A. Powell, G. Wilkinson, H. Mendez, J. Y. Ge, G. S. Huang, D. H. Miller, V. Pavlunin, I. P. J. Shipsey, B. Xin, G. S. Adams, D. Hu, B. Moziak, J. Napolitano, K. M. Ecklund, J. Insler, H. Muramatsu, C. S. Park, L. J. Pearson , et al. (56 additional authors not shown)

Abstract: Using the entire CLEO-c psi(3770) to DDbar event sample, corresponding to an integrated luminosity of 818 pb^-1 and approximately 5.4 x 10^6 DDbar events, we measure the form factors for the decays D0 to rho- e+ nu_e and D+ to rho0 e+ nu_e for the first time and the branching fractions with improved precision. A four-dimensional unbinned maximum likelihood fit determines the form factor ratios to… ▽ More Using the entire CLEO-c psi(3770) to DDbar event sample, corresponding to an integrated luminosity of 818 pb^-1 and approximately 5.4 x 10^6 DDbar events, we measure the form factors for the decays D0 to rho- e+ nu_e and D+ to rho0 e+ nu_e for the first time and the branching fractions with improved precision. A four-dimensional unbinned maximum likelihood fit determines the form factor ratios to be: V(0)/A_1(0) = 1.48 +- 0.15 +- 0.05 and A_2(0)/A_1(0)= 0.83 +- 0.11 +- 0.04. Assuming CKM unitarity, the known D meson lifetimes and our measured branching fractions we obtain the form factor normalizations A_1(0), A_2(0), and V(0). We also present a measurement of the branching fraction for D^+ to omega e^+ nu_e with improved precision. △ Less

Submitted 13 December, 2011; originally announced December 2011.

Comments: 9 pages, available through http://www.lns.cornell.edu/public/CLNS/, submitted to PRL

Report number: CLNS 11/2075; CLEO 11-3

Journal ref: Phys. Rev. Lett. 110, 131802 (2013)

Amplitude analyses of the decays chi_c1 -> eta pi+ pi- and chi_c1 -> eta' pi+ pi-

Authors: CLEO Collaboration, G. S. Adams, J. Napolitano, K. M. Ecklund, J. Insler, H. Muramatsu, C. S. Park, L. J. Pearson, E. H. Thorndike, S. Ricciardi, C. Thomas, M. Artuso, S. Blusk, R. Mountain, T. Skwarnicki, S. Stone, L. M. Zhang, G. Bonvicini, D. Cinabro, A. Lincoln, M. J. Smith, P. Zhou, J. Zhu, P. Naik, J. Rademacker , et al. (52 additional authors not shown)

Abstract: Using a data sample of 2.59 x 10^7 psi(2S) decays obtained with the CLEO-c detector, we perform amplitude analyses of the complementary decay chains chi_c1 -> eta pi+ pi- and chi_c1 -> eta' pi+ pi-. We find evidence for a P-wave eta' pi scattering amplitude, which, if interpreted as a resonance, would have exotic J^PC = 1^-+ and parameters consistent with the pi_1(1600) state reported in other pro… ▽ More Using a data sample of 2.59 x 10^7 psi(2S) decays obtained with the CLEO-c detector, we perform amplitude analyses of the complementary decay chains chi_c1 -> eta pi+ pi- and chi_c1 -> eta' pi+ pi-. We find evidence for a P-wave eta' pi scattering amplitude, which, if interpreted as a resonance, would have exotic J^PC = 1^-+ and parameters consistent with the pi_1(1600) state reported in other production mechanisms. We also make the first observation of the decay a_0(980) -> eta' pi and measure the ratio of branching fractions B(a_0(980) -> eta' pi)/B(a_0(980) -> eta pi) = 0.064 +- 0.014 +- 0.014. The pi pi spectrum produced with a recoiling eta is compared to that with eta' recoil. △ Less

Submitted 9 January, 2012; v1 submitted 27 September, 2011; originally announced September 2011.

Comments: 27 pages, 12 figures, available through http://www.lns.cornell.edu/public/CLNS/

Report number: CLNS 11/2080; CLEO 11-06

Journal ref: Phys. Rev. D 84, 112009 (2011)

Branching fractions for Y(3S) -> pi^0 h_b and psi(2S) -> pi^0 h_c

Authors: CLEO Collaboration, J. Y. Ge, D. H. Miller, I. P. J. Shipsey, B. Xin, G. S. Adams, J. Napolitano, K. M. Ecklund, J. Insler, H. Muramatsu, C. S. Park, L. J. Pearson, E. H. Thorndike, S. Ricciardi, C. Thomas, M. Artuso, S. Blusk, R. Mountain, T. Skwarnicki, S. Stone, L. M. Zhang, G. Bonvicini, D. Cinabro, A. Lincoln, M. J. Smith , et al. (51 additional authors not shown)

Abstract: Using e^+e^- collision data corresponding to 5.88M Y(3S) [25.9M psi(2S)] decays and acquired by the CLEO III [CLEO-c] detectors operating at CESR, we study the single-pion transitions from Y(3S) [psi(2S)] to the respective spin-singlet states h_{b[c]}. Utilizing only the momentum of suitably selected transition-pi^0 candidates, we obtain the upper limit B(Y(3S) -> pi^0 h_b) < 1.2\times 10^{-3} at… ▽ More Using e^+e^- collision data corresponding to 5.88M Y(3S) [25.9M psi(2S)] decays and acquired by the CLEO III [CLEO-c] detectors operating at CESR, we study the single-pion transitions from Y(3S) [psi(2S)] to the respective spin-singlet states h_{b[c]}. Utilizing only the momentum of suitably selected transition-pi^0 candidates, we obtain the upper limit B(Y(3S) -> pi^0 h_b) < 1.2\times 10^{-3} at 90% confidence level, and measure B(psi(2S) -> pi^0 h_c) = (9.0+-1.5+-1.3)\times 10^{-4}. Signal sensitivities are enhanced by excluding very asymmetric pi^0 -> gamma gamma candidates. △ Less

Submitted 22 August, 2011; v1 submitted 17 June, 2011; originally announced June 2011.

Comments: 12 pages 4 figures, version published in Physical Review D

Report number: CLNS 11/2079, CLEO 11-05

Journal ref: Phys.Rev.D84:032008,2011

Analysis of the Decay D^0 to K^0_S pi^0 pi^0

Authors: CLEO Collaboration, N. Lowrey, S. Mehrabyan, M. Selen, J. Wiss, J. Libby, M. Kornicer, R. E. Mitchell, M. R. Shepherd, C. M. Tarbert, D. Besson, T. K. Pedlar, J. Xavier, D. Cronin-Hennessy, J. Hietala, P. Zweber, S. Dobbs, Z. Metreveli, K. K. Seth, A. Tomaradze, T. Xiao, S. Brisbane, L. Martin, A. Powell, P. Spradlin , et al. (62 additional authors not shown)

Abstract: We present the results of a Dalitz plot analysis of D^0 to K^0_S pi^0 pi^0 using the CLEO-c data set of 818 inverse pico-barns of e^+ e^- collisions accumulated at sqrt{s} = 3.77 GeV. This corresponds to three million D^0 D^0-bar pairs from which we select 1,259 tagged candidates with a background of 7.5 +- 0.9 percent. Several models have been explored, all of which include the K^*(892), K^*_2(14… ▽ More We present the results of a Dalitz plot analysis of D^0 to K^0_S pi^0 pi^0 using the CLEO-c data set of 818 inverse pico-barns of e^+ e^- collisions accumulated at sqrt{s} = 3.77 GeV. This corresponds to three million D^0 D^0-bar pairs from which we select 1,259 tagged candidates with a background of 7.5 +- 0.9 percent. Several models have been explored, all of which include the K^*(892), K^*_2(1430), K^*(1680), the f_0(980), and the sigma(500). We find that the combined pi^0 pi^0 S-wave contribution to our preferred fit is (28.9 +- 6.3 +- 3.1)% of the total decay rate while D^0 to K^*(892)^0 pi^0 contributes (65.6 +- 5.3 +- 2.5)%. Using three tag modes and correcting for quantum correlations we measure the D^0 to K^0_S pi^0 pi^0 branching fraction to be (1.059 +- 0.038 +- 0.061)%. △ Less

Submitted 15 June, 2011; originally announced June 2011.

Comments: 11 pages, 3 figures, available through http://www.lns.cornell.edu/public/CLNS/, submitted to PRD

Report number: CLNS 10/2069, CLEO 10-06

Search for the decay $D^+_{s}\toωe^+ν$

Authors: CLEO Collaboration, L. Martin, A. Powell, G. Wilkinson, J. Y. Ge, D. H. Miller, I. P. J. Shipsey, B. Xin, G. S. Adams, B. Moziak, J. Napolitano, K. M. Ecklund, J. Insler, H. Muramatsu, C. S. Park, L. J. Pearson, E. H. Thorndike, S. Ricciardi, C. Thomas, M. Artuso, S. Blusk, R. Mountain, T. Skwarnicki, S. Stone, L. M. Zhang , et al. (51 additional authors not shown)

Abstract: We present the first search for the decay $D^+_{s}\to ωe^{+}ν$ to test the four-quark content of the $D^+_{s}$ and the $ω$-$φ$ mixing model for this decay. We use 586 $\mathrm{pb}^{-1}$ of $e^{+}e^{-}$ collision data collected at a center-of-mass energy of 4170 MeV. We find no evidence of a signal, and set an upper limit on the branching fraction of $\mathcal{B}(D^+_{s}\toωe^+ν)<$0.20% at the 90%… ▽ More We present the first search for the decay $D^+_{s}\to ωe^{+}ν$ to test the four-quark content of the $D^+_{s}$ and the $ω$-$φ$ mixing model for this decay. We use 586 $\mathrm{pb}^{-1}$ of $e^{+}e^{-}$ collision data collected at a center-of-mass energy of 4170 MeV. We find no evidence of a signal, and set an upper limit on the branching fraction of $\mathcal{B}(D^+_{s}\toωe^+ν)<$0.20% at the 90% confidence level. △ Less

Submitted 13 May, 2011; originally announced May 2011.

Comments: 18 pages, 9 figures

Report number: CLNS-11-2077

Journal ref: Phys.Rev.D84:012005,2011

Observation of the Dalitz Decay $D_{s}^{*+} \to D_{s}^{+} e^{+} e^{-}$

Authors: D. Cronin-Hennessy, J. Hietala, S. Dobbs, Z. Metreveli, K. K. Seth, A. Tomaradze, T. Xiao, L. Martin, A. Powell, G. Wilkinson, H. Mendez, J. Y. Ge, D. H. Miller, I. P. J. Shipsey, B. Xin, G. S. Adams, D. Hu, B. Moziak, J. Napolitano, K. M. Ecklund, J. Insler, H. Muramatsu, C. S. Park, L. J. Pearson, E. H. Thorndike , et al. (54 additional authors not shown)

Abstract: Using 586 $\textrm{pb}^{-1}$ of $e^{+}e^{-}$ collision data acquired at $\sqrt{s}=4.170$ GeV with the CLEO-c detector at the Cornell Electron Storage Ring, we report the first observation of $D_{s}^{*+} \to D_{s}^{+} e^{+} e^{-}$ with a significance of $5.3 σ$. The ratio of branching fractions $\calB(D_{s}^{*+} \to D_{s}^{+} e^{+} e^{-}) / \calB(D_{s}^{*+} \to D_{s}^{+} γ)$ is measured to be… ▽ More Using 586 $\textrm{pb}^{-1}$ of $e^{+}e^{-}$ collision data acquired at $\sqrt{s}=4.170$ GeV with the CLEO-c detector at the Cornell Electron Storage Ring, we report the first observation of $D_{s}^{*+} \to D_{s}^{+} e^{+} e^{-}$ with a significance of $5.3 σ$. The ratio of branching fractions $\calB(D_{s}^{*+} \to D_{s}^{+} e^{+} e^{-}) / \calB(D_{s}^{*+} \to D_{s}^{+} γ)$ is measured to be $[ 0.72^{+0.15}_{-0.13} (\textrm{stat}) \pm 0.10 (\textrm{syst})]%$, which is consistent with theoretical expectations. △ Less

Submitted 16 April, 2011; originally announced April 2011.

Observation of the h_c(1P) using e^+e^- collisions above DDbar threshold

Authors: CLEO Collaboration, T. K. Pedlar, D. Cronin-Hennessy, J. Hietala, S. Dobbs, Z. Metreveli, K. K. Seth, A. Tomaradze, T. Xiao, L. Martin, A. Powell, G. Wilkinson, H. Mendez, J. Y. Ge, D. H. Miller, I. P. J. Shipsey, B. Xin, G. S. Adams, D. Hu, B. Moziak, J. Napolitano, K. M. Ecklund, J. Insler, H. Muramatsu, C. S. Park , et al. (55 additional authors not shown)

Abstract: Using 586pb^-1 of e^+e^- collision data at E_CM = 4170MeV, produced at the CESR collider and collected with the CLEO-c detector, we observe the process e^+e^- --> pi^+ pi^- h_c(1P). We measure its cross section to be 15.6+-2.3+-1.9+-3.0pb, where the third error is due to the external uncertainty on the branching fraction of psi(2S) --> pi^0 h_c(1P), which we use for normalization. We also find evi… ▽ More Using 586pb^-1 of e^+e^- collision data at E_CM = 4170MeV, produced at the CESR collider and collected with the CLEO-c detector, we observe the process e^+e^- --> pi^+ pi^- h_c(1P). We measure its cross section to be 15.6+-2.3+-1.9+-3.0pb, where the third error is due to the external uncertainty on the branching fraction of psi(2S) --> pi^0 h_c(1P), which we use for normalization. We also find evidence for e^+e^- --> eta h_c(1P) at 4170MeV at the 3sigma level, and see hints of a rise in the e^+e^- --> pi^+ pi^- h_c(1P) cross section at 4260MeV. △ Less

Submitted 30 August, 2011; v1 submitted 11 April, 2011; originally announced April 2011.

Comments: 9 pages, 4 figures, as published

Journal ref: Phys.Rev.Lett.107:041803,2011

Semi-Inclusive Charged-Pion Electroproduction off Protons and Deuterons: Cross Sections, Ratios and Access to the Quark-Parton Model at Low Energies

Authors: R. Asaturyan, R. Ent, H. Mkrtchyan, T. Navasardyan, V. Tadevosyan, G. S. Adams, A. Ahmidouch, T. Angelescu, J. Arrington, A. Asaturyan, O. K. Baker, N. Benmouna, C. Bertoncini, H. P. Blok, W. U. Boeglin, P. E. Bosted, H. Breuer, M. E. Christy, S. H. Connell, Y. Cui, M. M. Dalton, S. Danagoulian, D. Day, J. A. Dunne, D. Dutta , et al. (55 additional authors not shown)

Abstract: A large set of cross sections for semi-inclusive electroproduction of charged pions ($π^\pm$) from both proton and deuteron targets was measured. The data are in the deep-inelastic scattering region with invariant mass squared $W^2$ > 4 GeV$^2$ and range in four-momentum transfer squared $2 < Q^2 < 4$ (GeV/c)$^2$, and cover a range in the Bjorken scaling variable 0.2 < x < 0.6. The fractional ener… ▽ More A large set of cross sections for semi-inclusive electroproduction of charged pions ($π^\pm$) from both proton and deuteron targets was measured. The data are in the deep-inelastic scattering region with invariant mass squared $W^2$ > 4 GeV$^2$ and range in four-momentum transfer squared $2 < Q^2 < 4$ (GeV/c)$^2$, and cover a range in the Bjorken scaling variable 0.2 < x < 0.6. The fractional energy of the pions spans a range 0.3 < z < 1, with small transverse momenta with respect to the virtual-photon direction, $P_t^2 < 0.2$ (GeV/c)$^2$. The invariant mass that goes undetected, $M_x$ or W', is in the nucleon resonance region, W' < 2 GeV. The new data conclusively show the onset of quark-hadron duality in this process, and the relation of this phenomenon to the high-energy factorization ansatz of electron-quark scattering and subsequent quark --> pion production mechanisms. The x, z and $P_t^2$ dependences of several ratios (the ratios of favored-unfavored fragmentation functions, charged pion ratios, deuteron-hydrogen and aluminum-deuteron ratios for $π^+$ and $π^-$) have been studied. The ratios are found to be in good agreement with expectations based upon a high-energy quark-parton model description. We find the azimuthal dependences to be small, as compared to exclusive pion electroproduction, and consistent with theoretical expectations based on tree-level factorization in terms of transverse-momentum-dependent parton distribution and fragmentation functions. In the context of a simple model, the initial transverse momenta of $d$ quarks are found to be slightly smaller than for $u$ quarks, while the transverse momentum width of the favored fragmentation function is about the same as for the unfavored one, and both fragmentation widths are larger than the quark widths. △ Less

Submitted 15 December, 2011; v1 submitted 8 March, 2011; originally announced March 2011.

Upsilon(1S)->gamma+f2'(1525); f2'(1525)->K0sK0s decays

Authors: The CLEO Collaboration, D. Besson, D. P. Hogan, T. K. Pedlar, D. Cronin-Hennessy, J. Hietala, P. Zweber, S. Dobbs, Z. Metreveli, K. K. Seth, A. Tomaradze, T. Xiao, S. Brisbane, L. Martin, A. Powell, P. Spradlin, G. Wilkinson, H. Mendez, J. Y. Ge, D. H. Miller, I. P. J. Shipsey, B. Xin, G. S. Adams, D. Hu, B. Moziak , et al. (61 additional authors not shown)

Abstract: We report on a study of exclusive radiative decays of the Upsilon(1S) resonance into a final state consisting of a photon and two K0s candidates. We find evidence for a signal for Upsilon(1S)->gamma f_2'(1525); f_2'(1525)->K0sK0s, at a rate (4.0+/-1.3+/-0.6)x10^{-5}, consistent with previous observations of Upsilon(1S)->gamma f_2'(1525); f_2'(1525)->K+K-, and isospin. Combining this branching frac… ▽ More We report on a study of exclusive radiative decays of the Upsilon(1S) resonance into a final state consisting of a photon and two K0s candidates. We find evidence for a signal for Upsilon(1S)->gamma f_2'(1525); f_2'(1525)->K0sK0s, at a rate (4.0+/-1.3+/-0.6)x10^{-5}, consistent with previous observations of Upsilon(1S)->gamma f_2'(1525); f_2'(1525)->K+K-, and isospin. Combining this branching fraction with existing branching fraction measurements of Upsilon(1S)->gamma f_2'(1525) and J/psi->gamma f_2'(1525), we obtain the ratio of branching fractions: B(Upsilon(1S)->gamma f_2'(1525))/B(J/psi->gamma f_2'(1525))=0.09+/-0.02, approximately consistent with expectations based on soft collinear effective theory. △ Less

Submitted 30 December, 2010; originally announced January 2011.

Comments: submitted to Brief Reports

Measurements of branching fractions for electromagnetic transitions involving the $χ_{bJ}(1P)$ states

Authors: The CLEO Collaboration, M. Kornicer, R. E. Mitchell, C. M. Tarbert, D. Besson, T. K. Pedlar, D. Cronin-Hennessy, J. Hietala, P. Zweber, S. Dobbs, Z. Metreveli, K. K. Seth, A. Tomaradze, T. Xiao, S. Brisbane, L. Martin, A. Powell, P. Spradlin, G. Wilkinson, H. Mendez, J. Y. Ge, D. H. Miller, I. P. J. Shipsey, B. Xin, G. S. Adams , et al. (60 additional authors not shown)

Abstract: Using 9.32, 5.88 million Upsilon(2S,3S) decays taken with the CLEO-III detector, we obtain five product branching fractions for the exclusive processes Upsilon(2S) => gamma chi_{b0,1,2}(1P) => gamma gamma Upsilon(1S) and Upsilon(3S) => gamma chi_{b1,2}(1P) => gamma gamma Upsilon(1S). We observe the transition chi_{b0}(1P) => gamma Upsilon(1S) for the first time. Using the known branching fractions… ▽ More Using 9.32, 5.88 million Upsilon(2S,3S) decays taken with the CLEO-III detector, we obtain five product branching fractions for the exclusive processes Upsilon(2S) => gamma chi_{b0,1,2}(1P) => gamma gamma Upsilon(1S) and Upsilon(3S) => gamma chi_{b1,2}(1P) => gamma gamma Upsilon(1S). We observe the transition chi_{b0}(1P) => gamma Upsilon(1S) for the first time. Using the known branching fractions for B[Upsilon(2S) => gamma chi_{bJ}(1P)], we extract values for B[chi_{bJ}(1P) => gamma Upsilon(1S)] for J=0, 1, 2. In turn, these values can be used to unfold the Upsilon(3S) product branching fractions to obtain values for B[Upsilon(3S) => gamma chi_{b1,2}(1P) for the first time individually. Comparison of these with each other and with the branching fraction B[Upsilon(3S) => gamma chi_{b0}] previously measured by CLEO provides tests of relativistic corrections to electric dipole matrix elements. △ Less

Submitted 16 March, 2011; v1 submitted 2 December, 2010; originally announced December 2010.

Comments: 22 pages, 6 figures, version published in Physical Review D

Report number: CLNS 10/2071, CLEO 10-08

Journal ref: Phys.Rev.D83:054003,2011