Sifter: Scalable Sampling for Distributed Traces, without Feature Engineering
Open accessThe DOI matches its Crossref record; 32 of the 35 references it cites were found in a registry or an index.
Abstract
Distributed tracing is a core component of cloud and datacenter systems, and provides visibility into their end-to-end runtime behavior. To reduce computational and storage overheads, most tracing frameworks do not keep all traces, but sample them uniformly at random. While effective at reducing overheads, uniform random sampling inevitably captures redundant, common-case execution traces, which are less useful for analysis and troubleshooting tasks. In this work we present Sifter, a general-purpose framework for biased trace sampling. Sifter captures qualitatively more diverse traces, by weighting sampling decisions towards edge-case code paths, infrequent request types, and anomalous events. Sifter does so by using the incoming stream of traces to build an unbiased low-dimensional model that approximates the system's common-case behavior. Sifter then biases sampling decisions towards traces that are poorly captured by this model. We have implemented Sifter, integrated it with several open-source tracing systems, and evaluate with traces from a range of open-source and production distributed systems. Our evaluation shows that Sifter effectively biases towards anomalous and outlier executions, is robust to noisy and heterogeneous traces, is efficient and scalable, and adapts to changes in workloads over time.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Figures















Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“high storage and runtime costs demand sampling strategies, where random head-based or selective tail-based sampling reduce overheads while preserving representative traces [75]”
From Anomaly detection and root-cause identification in microservices: a survey (2026) · cites this paper as [75]
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Time, clocks, and the ordering of events in a distributed system
Lamport, L. (1978) · Communications of the ACM · doi:10.1145/359545.359563 · Cited by 8,488 · cited 1×
“Relationships Events alone tell some of the story, but the most important information in a trace is the causal ordering between events, defined by Lamport’s happened-before relation [16].”
BackgroundDistributed Representations of Sentences and Documents
Le, Q. V., & Mikolov, T. (2014) · arXiv (Cornell University) · doi:10.48550/arxiv.1405.4053 · Cited by 4,934 · cited 6×
“We implement the model using the same architecture as the distributed memory paragraph vector model, using concatenation [18].”
BackgroundMethodInfluentialResilient distributed datasets: a fault-tolerant abstraction for in-memory cluster computing
Zaharia, M., Chowdhury, M., Das, T., Dave, A., Ma, J. H., McCauley, M., Franklin, M. J. M., Shenker, S. J., & Stoica, I. G. (2012) · UC Berkeley · Cited by 4,891
The Hadoop Distributed File System
Shvachko, K., Kuang, H., Radia, S., & Chansler, R. (2010) · 2010 IEEE 26th Symposium on Mass Storage Systems and Technologies (MSST) · doi:10.1109/msst.2010.5496972 · Cited by 4,834
A Neural Probabilistic Language Model
Bengio, Y., Ducharme, R., Vincent, P., Jauvin, C., Jaz, K., Hofmann, T., Poggio, T., & Shawe-Taylor, J. S. (2000) · Cited by 3,486
Apache Hadoop YARN: yet another resource negotiator
Vavilapalli, V. K., Murthy, A. C., Douglas, C., Agarwal, S., Konar, M., Evans, R., Graves, T., Lowe, J., Shah, H., Seth, S., et al. (2013) · Proceedings of the 4th annual Symposium on Cloud Computing · doi:10.1145/2523616.2523633 · Cited by 1,848
Detecting large-scale system problems by mining console logs
Xu, W., Huang, L., Fox, A., Patterson, D., & Jordan, M. I. (2009) · Proceedings of the ACM SIGOPS 22nd symposium on Operating systems principles · doi:10.1145/1629575.1629587 · Cited by 1,332
Vishwanathan, S. V. N., Schraudolph, N. N., Kondor, R., & Borgwardt, K. M. (2010) · CaltechAUTHORS (California Institute of Technology) · Cited by 1,305
Dapper, a Large-Scale Distributed Systems Tracing Infrastructure
Sigelman, B. H., Barroso, L. A., Burrows, M., Stephenson, P., Plakal, M., Beaver, D., Jaspan, S., & Shanbhag, C. (2010) · Cited by 899
Gan, Y., Zhang, Y., Cheng, D., Shetty, A., Rathi, P., Katarki, N., Bruno, A., Hu, J., Ritchken, B., Jackson, B., et al. (2019) · Proceedings of the Twenty-Fourth International Conference on Architectural Support for Programming Languages and Operating Systems · doi:10.1145/3297858.3304013 · Cited by 665
Research works printed but not linked: 8
- M. P. Kasick, J. Tan, R. Gandhi, and P. Narasimhan. Black-Box Problem Diagnosis in Parallel File Systems. In 8th USENIX Conference on File and Storage Technologies (FAST '10). ( §9).
- J. Leavitt. End-to-End Tracing Models: Analysis and Unification. B.Sc. Thesis, Brown University, 2014. ( §4.1).
- J. Mace. Revisiting End-to-End Trace Comparison with Graph Kernels. M.Sc. Project, Brown University, 2013. ( §3.2, 4.3, and 9).
- G. Mann, M. Sandler, D. Krushevskaja, S. Guha, and E. Even-Dar. Modeling the Parallel Execution of Black-Box Services. In 3rd USENIX Workshop on Hot Topics in Cloud Computing (HotCloud '11). ( §4.3, 4.4, and 9).
- T. Mikolov, I. Sutskever, K. Chen, G. S. Corrado, and J. Dean. Distributed Repre- sentations of Words and Phrases and their Compositionality. In 27th Conference on Neural Information Processing Systems (NIPS '13). ( §5.1 and 9).
- K. Nagaraj, C. Killian, and J. Neville. Structured Comparative Analysis of System Logs to Diagnose Performance Problems. In 9th USENIX Conference on Networked Systems Design and Implementation (NSDI '12). ( §4.2 and 9).
- K. Ostrowski, G. Mann, and M. Sandler. Diagnosing Latency in Multi-Tier Black- Box Services. In 5th Workshop on Large Scale Distributed Systems and Middleware (LADIS '11). ( §9).
- H. Tian, M. Yu, and W. Wang. Continuum: A Platform for Cost-Aware, Low- Latency Continual Learning. In 10th ACM Symposium on Cloud Computing (SOCC '18). ( §9).
Cited by
Network-Centric Distributed Tracing with DeepFlow: Troubleshooting Your Microservices in Zero Code
Shen, J., Zhang, H., Xiang, Y., Shi, X., Li, X., Shen, Y., Zhang, Z., Wu, Y., Yin, X., Wang, J., et al. (2023) · ACM SIGCOMM Conference (SIGCOMM) · doi:10.1145/3603269.3604823 · Cited by 88
Automated Analysis of Distributed Tracing: Challenges and Research Directions
Bento, A., Correia, J., Filipe, R., Araujo, F., & Cardoso, J. (2021) · Journal of Grid Computing · doi:10.1007/s10723-021-09551-5 · Cited by 45
“approach in our suggestions to improve the usefulness of tracing for automated analysis—[25] propose Sifter, a trace sampler built to bias the sampling decision towards edge cases and rare work-flows.”
MethodHuang, L., & Zhu, T. (2021) · Proceedings of the ACM Symposium on Cloud Computing · doi:10.1145/3472883.3486994 · Cited by 37
TraceCRL: contrastive representation learning for microservice trace analysis
Zhang, C., Peng, X., Zhou, T., Sha, C., Yan, Z., Chen, Y., & Yang, H. (2022) · Proceedings of the 30th ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering · doi:10.1145/3540250.3549146 · Cited by 36
“The same authors [11] propose a neural network-based trace sampling approach which constructs trace vectors based on fixed-length operation sequences.”
BackgroundCloud Data Services: Workloads, Architectures and Multi-Tenancy
Narasayya, V. R., & Chaudhuri, S. (2021) · Foundations and Trends in Databases · doi:10.1561/9781680837759 · Cited by 38
Multi-Tenant Cloud Data Services: State-of-the-Art, Challenges and Opportunities
Narasayya, V. R., & Chaudhuri, S. (2022) · ACM SIGMOD International Conference on Management of Data (SIGMOD) · doi:10.1145/3514221.3522566 · Cited by 38
Open tracing tools: Overview and critical comparison
Janes, A., Li, X., & Lenarduzzi, V. (2023) · Journal of Systems and Software · doi:10.1016/j.jss.2023.111793 · Cited by 28
Sieve: Attention-based Sampling of End-to-End Trace Data in Distributed Microservice Systems
Huang, Z., Chen, P., Yu, G., Chen, H., & Zheng, Z. (2021) · 2021 IEEE International Conference on Web Services (ICWS) · doi:10.1109/icws53863.2021.00063 · Cited by 22
Tracemesh: Scalable and Streaming Sampling for Distributed Traces
Chen, Z., Jiang, Z., Su, Y., Lyu, M. R.-T., & Zheng, Z. (2024) · IEEE ... International Conference on Cloud Computing · doi:10.1109/cloud62652.2024.00016 · Cited by 21
“Repre-sentative approaches are introduced in Section IV-A3, which leverage a variety of methods, including tree-based models [10], [11], clustering algorithms [8], and neural language techniques [9].”
BackgroundMethodInfluentialInvestigating Performance Overhead of Distributed Tracing in Microservices and Serverless Systems
Nõu, A., Talluri, S., Iosup, A., & Bonetta, D. (2025) · ACM/SPEC International Conference on Performance Engineering (ICPE) · doi:10.1145/3680256.3721316 · Cited by 24
LatenSeer: Causal Modeling of End-to-End Latency Distributions by Harnessing Distributed Tracing
Zhang, Y., Isaacs, R., Yue, Y., Yang, J., Zhang, L., & Vigfússon, Ý. (2023) · ACM Symposium on Cloud Computing (SoCC) · doi:10.1145/3620678.3624787 · Cited by 22
“The primary use cases of distributed end-to-end tracing systems concern active system monitoring [8, 57, 72], anomaly detection [6, 11, 27, 39, 47, 65, 69, 73], and root-cause analysis [55, 83].”
BackgroundUsing Distributed Tracing to Identify Inefficient Resources Composition in Cloud Applications
Cassé, C., Berthou, P., Owezarski, P., & Josset, S. (2021) · IEEE International Conference on Cloud Networking · doi:10.1109/cloudnet53349.2021.9657140 · Cited by 22
STEAM: Observability-Preserving Trace Sampling
He, S., Feng, B., Li, L., Zhang, X., Kang, Y., Lin, Q., Rajmohan, S., & Zhang, D. (2023) · ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering (ESEC/FSE) · doi:10.1145/3611643.3613881 · Cited by 21
Mertz, J., & Nunes, I. (2022) · Journal of Systems and Software · doi:10.1016/j.jss.2023.111708 · Cited by 14
“Similarly, Las-Casas [13, 34] aim at capturing outliers and anomalous traces.”
BackgroundTraStrainer: Adaptive Sampling for Distributed Traces with System Runtime State
Huang, H., Zhang, X., Chen, P., He, Z., Zhi-ming, C., Yu, G., Chen, H., & Sun, C. (2024) · Proceedings of the ACM on software engineering. · doi:10.1145/3643748 · Cited by 18
3MileBeach: A Tracer with Teeth
Zhang, J., Ferydouni, R., Montana, A., Bittman, D., & Alvaro, P. (2021) · ACM Symposium on Cloud Computing (SoCC) · doi:10.1145/3472883.3486986 · Cited by 13
Aggregate-Driven Trace Visualizations for Performance DebuggingPreprint
Anand, V., Stolet, M., Davidson, T., Beschastnikh, I., Munzner, T., & Mace, J. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2010.13681 · Cited by 16
“Early work on trace comparison similarly compared latency distributions, but required traces to be structurally isomorphic [22,23]; this is rare in practice as most traces are structurally unique [17].”
Self-citationTracezip: Efficient Distributed Tracing via Trace Compression
Chen, Z., Pu, J., & Zheng, Z. (2025) · Proceedings of the ACM on software engineering. · doi:10.1145/3728888 · Cited by 14
“Panorama [16] achieves more sophisticated observability by collecting inter-process and inter-thread error signals.”
Cotroneo, D., Simone, L. D., Liguori, P., & Natella, R. (2023) · Journal of Systems and Software · doi:10.1016/j.jss.2023.111611 · Cited by 11
Blueprint: A Toolchain for Highly-Reconfigurable Microservice Applications
Anand, V., Garg, D., Kaufmann, A., & Mace, J. (2023) · ACM SIGOPS Symposium on Operating Systems Principles (SOSP) · doi:10.1145/3600006.3613138 · Cited by 11
Self-citationMint: Cost-Efficient Tracing with All Requests Collection via Commonality and Variability Analysis
Huang, H., Chen, C., Chen, K., Chen, P., Yu, G., He, Z., Wang, Y., Zhang, H., & Zhou, Q. (2025) · International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS) · doi:10.1145/3669940.3707287 · Cited by 14
“This means that prior methods only reduce the number of traces without lightweighting each trace itself [25, 29, 52].”
InfluentialTracePicker: Optimization-Based Trace Sampling for Microservice-Based Systems
Xie, S., Wang, J., Li, M., Chen, P., Xuan, J., & Li, B. (2025) · Proceedings of the ACM on software engineering. · doi:10.1145/3729351 · Cited by 9
“However, existing samplers, especially structural-based sampling methods [15], [16], can hardly identify anomalous traces, primarily due to the neglect of temporal or status information (see § 2.2.1).”
InfluentialOpen Tracing Tools: Overview and Critical ComparisonPreprint
Janes, A., Li, X., & Lenarduzzi, V. (2022) · SSRN Electronic Journal · doi:10.2139/ssrn.4175937 · Cited by 7
Distributed computation of the critical path from execution traces
Denys, P., Fournier, Q., & Dagenais, M. (2023) · Software Practice and Experience · doi:10.1002/spe.3210 · Cited by 6
A Survey of Program Analysis for Distributed Software Systems
Cai, H. (2025) · ACM Computing Surveys · doi:10.1145/3742900 · Cited by 6
Snicket: Query-Driven Distributed Tracing
Berg, J., Ruffy, F., Nguyen, K., Yang, N., Kim, T., Sivaraman, A., Netravali, R., & Narayana, S. (2021) · ACM Workshop on Hot Topics in Networks (HotNets) · doi:10.1145/3484266.3487393 · Cited by 12
SampleHST: Efficient On-the-Fly Selection of Distributed Traces
Gias, A. U., Gao, Y., Sheldon, M., Perusquía, J. A., O’Brien, O., & Casale, G. (2023) · IEEE/IFIP Network Operations and Management Symposium (NOMS) · doi:10.1109/noms56928.2023.10154383 · Cited by 6
Anomaly detection and root-cause identification in microservices: a survey
Barata, L. M., Sequeira, S., Lopes, E., Inácio, P. R. M., & Freire, M. M. (2026) · Cluster Computing · doi:10.1007/s10586-026-06095-9 · Cited by 4
Unleashing Performance Insights with Online Probabilistic Tracing
Toslali, M., Qasim, S., Parthasarathy, S., Oliveira, F. A., Huang, H., Stringhini, G., Liu, Z., & Coskun, A. K. (2024) · IEEE International Conference on Cloud Engineering · doi:10.1109/ic2e61754.2024.00015 · Cited by 5
“Distributed tracing often employs request-based sampling as a means to mitigate costs, allowing for tracing only a subset of requests [23], [25], [30].”
InfluentialThe Diagnosis-Effective Sampling of Application Traces
Poghosyan, A., Harutyunyan, A. N., Davtyan, E., Petrosyan, K., & Baloian, N. A. (2024) · Applied Sciences · doi:10.3390/app14135779 · Cited by 4
Geng, Y., Zhang, H., Wu, Z., Li, Y., Wang, J., & Yin, X. (2025) · ACM SIGCOMM Conference (SIGCOMM) · doi:10.1145/3718958.3750477 · Cited by 6
AutoScope: Code Knowledge Enhanced Span-level Sampling for Distributed Tracing
Wu, Y., Yu, G., Jiang, Z., Li, Y., & Lyu, M. R.-T. (2026) · ACM Transactions on Software Engineering and Methodology · doi:10.1145/3830907 · Cited by 4
“To address these limitations, recent studies have explored biased sampling techniques, leveraging methods such as tree-based models [24], clustering algorithms [29], and neural language processing techniques [30].”
InfluentialThe case for an internet primitive for fault localization
Sussman, W., Marx, E., Arun, V., Narayan, A., Alizadeh, M., Balakrishnan, H., Panda, A., & Shenker, S. J. (2022) · ACM Workshop on Hot Topics in Networks (HotNets) · doi:10.1145/3563766.3564105 · Cited by 4
Search-Based Performance Testing and Analysis for Microservice-Based Digital Power Applications
Pei, Q., Liang, Z., Wang, Z., Cui, L., Long, Z., & Wu, G. (2023) · International Conference on Energy, Electrical and Power Engineering (CEEPE) · doi:10.1109/ceepe58418.2023.10165808 · Cited by 3
“Recent studies have shown that distributed tracing is a powerful tool for debugging [12] [13].”
BackgroundGias, A. U., Gao, Y., Sheldon, M., Perusquía, J. A., O’Brien, O., & Casale, G. (2024) · Journal of Network and Systems Management · doi:10.1007/s10922-024-09818-8 · Cited by 3
Metadata-Centric Orchestration for Cloud-Native ETL Pipelines
Verma, A. (2025) · International Journal of Computational and Experimental Science and Engineering · doi:10.22399/ijcesen.4292 · Cited by 1
LightTrace: A Versatile Ebpf-Enabled Toolkit for Lightweight Distributed Tracing
Zhang, Y., Wang, K. Y., Gong, S., Liang, H., Hu, C., & Zhou, X. (2025) · IEEE International Conference on Parallel and Distributed Systems (ICPADS) · doi:10.1109/icpads67057.2025.11322958 · Cited by 1
Shahedi, K., Khouzam, M., Li, H., Lamothe, M., & Khomh, F. (2025) · IEEE/ACM International Conference on Automated Software Engineering (ASE) · doi:10.1109/ase63991.2025.00286 · Cited by 2
Scommegna, L., Picano, B., Verdecchia, R., & Vicario, E. (2025) · Journal of Systems and Software · doi:10.1016/j.jss.2025.112518 · Cited by 1
Estimation from Partially Sampled Distributed TracesPreprint
Ertl, O. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2107.07703 · Cited by 1
“2 Head-Based Trace Sampling Head-based sampling [7, 8, 12, 13], also called up-front sampling [11], is a simple and widely used sampling approach for traces.”
MethodMicrosserviços aplicados no gerenciamento de dados de vistorias imobiliárias: um estudo de caso
Müller, R. H., Meinhardt, C., & Mendizabal, O. M. (2020) · doi:10.5753/wcga.2020.12443
“Isso pode ser obtido gerando um identificador externo único, que é colocado em todos os logs [Richardson 2019, Las-Casas et al. 2019].”
BackgroundBee: End to End Distributed Tracing System for Source Code Security Analysis
Qiu, L., Song, X., Yang, J. J., & Cui, B. (2022) · Highlights in Science Engineering and Technology · doi:10.54097/hset.v1i.463 · Cited by 1
Tracing Sites Requests in Using Web Applications
Ihsanova, K., Rezedinova, E., & Sergeev, A. (2023) · Lecture notes in networks and systems · doi:10.1007/978-3-031-20875-1_31
Dynamic Load Scaling via Distributed Trace Sampling and Analysis
Mooney, K., Durkin, J., Koshy, S., Khadka, S., Gu, J. T., & Farivar, R. (2025) · IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing (CCGrid) · doi:10.1109/ccgrid64434.2025.00050
Machine Learning Applications in Databases and Big Data Analytics
Dritsas, E., Τρίγκα, Μ., & Phivos Mylonas, B. (2025) · Lecture notes in networks and systems · doi:10.1007/978-3-032-06634-3_8
Distributed Performance Analysis Tools for Large Scale Computations
Denys, P.-F. (2023) · PolyPublie (École Polytechnique de Montréal) · Cited by 1
SLOTH: Lightweight Detection and Localization of On-Chip Fail-Slow Failures for DNN AcceleratorsPreprint
Wu, J., Wan, X., Li, Z., Jin, Y., Sun, G., Liang, Y. E., Zhou, D., & Zhuo, Y. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2510.24112
“Existing distributed tracing systems reduce overhead through reactive collection or selective storage[29], [33], [32], [58], but these approaches are often not suitable for diagnosing fail-slow behaviors.”
Ramverk för svansbaserad provtagning för distribuerad spårning med hjälp av strömbearbetning
Shuvo, K. (2021) · Aaltodoc (Aalto University)
Predicting the performance of Web services in a Cloud environment
Cassé, C. (2023) · theses.fr (ABES) · doi:10.70675/bb3116efz9360z40b7za7d5z2dd49fff7025
GuardGRPC: Priority-Aware Resource Governance for OTLP/gRPC Telemetry Export
Gullapalli, S., Divi, V. R., Muppidi, N. K. R., Rachamalla, S., & Tatavarthi, S. (2026) · International Conference on Electronics Representation and Algorithm (ICERA) · doi:10.1109/icera72709.2026.11666727 · Cited by 2
“Sifter prioritized unusual traces, Hind-sight retained traces retroactively after symptoms appeared, and OpenTelemetry tail sampling supports configurable policies based on error status, latency, or probabilistic decisions [9]–[11].”
Gleaner: A Semantically-Rich and Efficient Online Sampler for Microservice Diagnostics
Yang, Y., Fang, A., Zhang, S., & He, P. (2026) · Proceedings of the ACM on software engineering. · doi:10.1145/3832241
Metadata
- Authors
- Las-Casas, P., Papakerashvili, G., Anand, V., & Mace, J.
- Affiliations
- Max Planck Institute for Software Systems (DE)
- Venue
- Proceedings of the ACM Symposium on Cloud Computing
- Publisher
- ACM
- Published
- 2019-11-20
- Pages
- 312-324
- OpenAlex
- W2988883278
- Cited by
- 42
- References
- 35
- Citations by year
- 2026: 3 · 2025: 12 · 2024: 4 · 2023: 10 · 2022: 5 · 2021: 7 · 2020: 1
- FWCI
- 2.81
- Type
- conference paper
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- Locations
- ACM Symposium on Cloud Computing (SoCC) (publishedVersion, cc-by)
- Topics
- Physical Sciences › Computer Science › Computer Networks and Communications › Software System Performance and Reliability
- Physical Sciences › Computer Science › Information Systems › Cloud Computing and Resource Management
- Physical Sciences › Computer Science › Information Systems › Advanced Database Systems and Queries
- Keywords
- distributed tracing, uniform random sampling, anomaly detection, low-dimensional models, SIFTER
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.