Pith. sign in

REVIEW 4 major objections 6 minor 1 cited by

AdaDeDup: Adaptive Hybrid Data Pruning for Efficient Large-Scale Object Detection Training

T0 review · 4 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read AdaDeDup claims hybrid density-plus-model pruning trains object detectors on 20% less data with almost no accuracy loss, cutting performance loss by at least 54% versus random downsampling on Waymo.

desk verdict The adaptive signal in Eq. (2) is confounded by pruning ratio and cluster size, so the paper's central mechanism is not supported, but the empirical study is solid enough to warrant a careful revision. read the letter →

arxiv 2507.00049 v1 pith:TLKD7UN6 submitted 2025-06-24 cs.CV cs.LG

Feiyang Kang , Nadine Chang , Maying Shen , Marc T. Law , Rafid Mahmood , Ruoxi Jia , Jose M. Alvarez This is my paper
classification cs.CVcs.LG
keywords datapruningobjectdetectiondensity-basedproxymodelcluster-adaptiveselectionefficiencysemanticdeduplicationautonomousdriving
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper claims that the two competing families of data pruning can be merged into one adaptive rule that works for large-scale object detection. Density-based pruning removes near-duplicate images cheaply but is blind to what the detector actually needs, while model-based scoring tracks the task but is costly and can over-select hard or noisy samples. AdaDeDup first prunes inside semantic clusters, then trains a small proxy model on the kept images and measures, per cluster, whether the pruned images were easier or harder than the kept ones using the differential loss signal $\Delta\ell_i$. Clusters whose pruned images were easy get pruned harder; clusters whose pruned images were hard get some samples restored. On Waymo, COCO, and nuScenes the paper reports near-original mean average precision at 15–20% pruning and reductions in performance loss of at least 54% (Waymo) and 66% (COCO) relative to random downsampling, so the practical stake is training accurate detectors on noticeably less data and compute.

What carries the argument

The carrying object is the differential loss signal $\Delta\ell_i = \ell_i^s - \ell_i^p$ (Eq. 2): for each cluster $c_i$, the proxy model's summed loss on initially kept samples minus its summed loss on initially pruned samples. The paper treats this scalar as a zero-order estimate of the gradient of the selection objective with respect to the cluster's pruning parameter, converting an NP-hard bi-level subset-selection problem into one threshold update $\gamma'_i = \mathrm{clip}(\gamma_i + \beta \tilde{\Delta\ell}_i, 0, 1)$, with the per-cluster updates normalized so the total data budget is preserved. The sign of $\Delta\ell_i$ decides whether cluster $c_i$ is pruned more or less; its magnitude, scaled by $\alpha_+$ or $\alpha_-$, sets how far the cluster-specific density threshold moves in that single step.

What would settle it

A decisive check: train the proxy on the initially kept subset, then retrain a second proxy on the kept subset plus a random half of the pruned samples, and see whether the per-cluster loss-gap ranking of which clusters deserve less pruning survives. If the ranking flips for many clusters, or if AdaDeDup's mAP gain over its density baseline disappears when the swapped-back 5–10% of samples are chosen randomly instead of by the loss gap, the adaptive signal is not the cause of the reported results.

Watch

Extended reading notes

Core claim

The central claim is that a signed per-cluster loss gap, $\Delta\ell_i = \ell_i^s - \ell_i^p$, computed by a proxy model trained only on the initially kept subset, is a usable signal for how aggressively each semantic cluster should be pruned. If within a cluster the kept samples carry higher average loss than the pruned ones, the pruned ones are judged redundant and the cluster's pruning ratio rises; if the pruned samples carry higher loss, informative content was discarded and the ratio falls. The adjustment is a single one-shot update that changes the selection status of roughly 5–10% of samples, avoiding the repeated retraining of bi-level optimization. The paper argues that this cluster-adaptive hybrid beats pure density-based deduplication (CLIP-DeDup and VLM-SSE) and random downsampling on Waymo, COCO, and nuScenes, reaching near-full-data mAP at 20% pruning on Waymo and 10% on COCO, and matching random downsampling's accuracy with 15–20% less data.

Load-bearing premise

The method assumes that the difference in a proxy model's loss between kept and pruned images inside a cluster tells you how much real information the pruning threw away, even though the proxy was trained only on the kept images and might simply be more comfortable with them.

Editorial extensions

If this is right

  • At 20% pruning on Waymo and 10% on COCO, models trained on the pruned subset reach nearly the same mean average precision as full-data training, so a fifth of a large detection dataset can be dropped without a retraining-quality penalty.
  • The adaptation is a single one-shot update that reallocates each cluster's pruning budget, so no line search or repeated retraining is needed; the paper reports choosing the update size so that roughly 5–10% of sample selections change.
  • Because results are reported with BEVFormer-S on driving benchmarks and Faster R-CNN on COCO, the mechanism appears to transfer across scene-structured multi-camera data and generic image collections.
  • The largest relative gains over random downsampling appear at higher pruning ratios (at least 54% loss reduction on Waymo up to 40% pruning; at least 66% on COCO up to 20% pruning), so the method's advantage grows as the data budget tightens.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The per-cluster loss gap is a coarse influence estimate: it replaces expensive inverse-Hessian computations with one forward pass of a small proxy. The paper does not test whether the signal remains reliable when the proxy is far smaller than the target model, even though its own proxies already train on at most 10–30k samples.
  • The main untested risk is self-confirmation: because the proxy trains only on kept samples, a negative gap may reflect overfitting to those samples rather than true redundancy. A direct test would retrain the proxy on a mix of kept and pruned samples and check whether cluster-level decisions flip.
  • Waymo and nuScenes are scene-structured: removing one front-view image discards the whole multi-camera scene, so the effective pruning decision is per scene while the loss gap is computed per image; decoupling these two levels is a refinement the paper leaves open.
  • The paper lists sensitivity to cluster count and to the scaling constants $\alpha_+$, $\alpha_-$, $\beta$ as a limitation; normalizing the gap by cluster size or intra-cluster density variance is a testable way to stabilize the adaptation across 10, 100, and 500 clusters.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. AdaDeDup proposes a two-stage hybrid data-pruning method for object detection. Stage 1 clusters the dataset with VLM features and applies an initial density-based pruning. Stage 2 trains a proxy model on the kept subset, computes per-cluster differential loss signals by summing losses on kept versus pruned samples, and uses these signals to adjust each cluster's pruning ratio before re-pruning from the full dataset. Experiments on Waymo, COCO, and nuScenes with BEVFormer-S and Faster R-CNN report that AdaDeDup outperforms random downsampling, CLIP-DeDup, and VLM-SSE, achieving near-original mAP at 20% pruning on Waymo and reducing performance loss substantially relative to baselines.

Significance. If the proposed mechanism is sound, AdaDeDup is a practically relevant contribution: it targets the under-explored setting of data pruning for object detection, combines density-based and model-based signals at cluster level, and is evaluated on three large-scale benchmarks with standard deviations over multiple runs and open-sourced code. However, the central adaptive signal is confounded by cluster size and initial pruning ratio, so the current experiments do not establish the claimed mechanism. The method may still work as a heuristic, but the paper's theoretical framing and the attribution of the empirical gains need correction. The manuscript's own limitations section acknowledges sensitivity to hyperparameters but does not address this confounding.

major comments (4)
  1. [§3.3.2, Eq. (2) and Algorithm 1, lines 8-9] The differential signal Δℓ_i = ℓ_i^s − ℓ_i^p is computed as the difference of unnormalized sums over kept and pruned subsets that have different sizes. If the per-sample average loss in cluster c_i is μ, then Δℓ_i = μ |c_i| (1 − 2γ_i), where γ_i is the initial pruning ratio. Consequently, the sign of Δℓ_i is determined by whether γ_i is below or above 1/2, not by whether the pruned samples are redundant. Since the update γ'_i = γ_i + β·α_i·Δℓ_i increases γ_i when Δℓ_i > 0 and decreases it when Δℓ_i < 0, the signal drives every cluster's pruning ratio toward 1/2 regardless of model feedback. The reported experiments therefore do not identify the proposed model-informed adaptation as the source of the gains. The manuscript should compute ℓ_i^s and ℓ_i^p as per-sample averages (or otherwise normalize by subset size), rerun the experiments, and reinterpret the results.
  2. [§3.2, Step 2, and Algorithm 1] The claimed zero-order gradient approximation, ℓ(θ*(W_s), s_p) − ℓ(θ*(W_s), s_k) as a proxy for ∂J(W_s)/∂(W_s)_p, is not derived, and it is not what Algorithm 1 computes: Algorithm 1 compares aggregated losses over entire clusters rather than a kept sample near each pruned sample. Without a formal connection between the aggregate Δℓ_i and the policy gradient ∂J/∂λ_c, the theoretical framing in Section 3.2 is not supported by the implementation. Please either provide a rigorous derivation or state explicitly that the update is a heuristic motivated by, but not equivalent to, a zero-order gradient.
  3. [§3.3.2 and Algorithm 1, step 14] The normalization step is underspecified. The text says the scaled differences are 'typically normalized' and Algorithm 1 states that the adjusted values must satisfy ∑ |c_i|(γ_i + β·Δℓ_i*) = n − m while respecting 0 ≤ γ'_i ≤ 1, but no concrete normalization or rounding procedure is given. The final selected set depends on this normalization, and feasibility after clipping is not established. Specify the exact normalization rule, the order of clipping versus rescaling, and how rounding of k_i is handled.
  4. [§3.3.2, Step 1] Because the proxy model is trained only on the kept subset D_s^(0), the loss difference ℓ_i^s − ℓ_i^p reflects the model's familiarity with the kept samples as much as any intrinsic redundancy of the pruned samples. A proxy trained on a random subset of the same size would also produce nonzero per-cluster differences, so the adaptive signal may be dominated by overfitting to the kept set. The paper does not test this possibility. An ablation with a proxy trained on a random subset of equal size, or another control, is needed to support the claim that Δℓ_i measures relative information content.
minor comments (6)
  1. [Algorithm 1, lines 4-6] The proxy model training appears inside the first for loop, which would train the proxy K times; move it before the loop or fix the indentation.
  2. [Section 2 and Section 3.1] GLISTER is cited as [14], but reference [14] is GoodCore; the correct GLISTER citation appears to be [37]. Please fix the citation.
  3. [Section 3.2, item 2] The heading 'Parameterizing the pruning policy to narrow the decision space' is repeated from item 1; the heading for item 2 should be 'Zero-order estimation for the policy gradient'.
  4. [Appendix D.1] 'A V datasets' should be 'AV datasets'.
  5. [Appendix C.1] The GPU name 'NVIDIA A5880 Ada' appears to be a typo for 'NVIDIA A6000 Ada' or the correct product name.
  6. [Broader Impact Statement] The word 'perserved' should be 'preserved'.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity; the central claims are anchored by held-out benchmark evaluation, with only minor non-load-bearing self-citations.

full rationale

AdaDeDup's central claims are empirical: models trained on the selected subsets are evaluated on held-out validation splits, and the reported mAP numbers (Waymo, COCO, nuScenes) are not used as fitting targets in the selection algorithm. The differential-loss signal in Eq. (2) is explicitly framed as a heuristic zero-order proxy for marginal utility (Section 3.2), not as a quantity fitted to the evaluation metric and then renamed as a prediction. The authors do cite their own prior work, most notably VLM-SSE [5] as a baseline and as the source of the specialized AV prompt, plus several optimization and data-selection papers from the same group; these citations are not load-bearing for the main empirical claim, and no uniqueness theorem or external mathematical result is imported from the authors' own prior papers to force the method. The paper's limitation section honestly acknowledges proxy-model sensitivity and hyperparameter dependence. A legitimate reviewer concern is that Eq. (2) sums unnormalized losses over unequal-sized kept and pruned subsets per cluster, so if average losses are similar the sign of Δℓ_i is dominated by the initial per-cluster pruning ratio γ_i rather than by sample informativeness; this is a possible confound or identifiability issue in the mechanism's interpretation, but it does not make the held-out mAP comparison circular, because the final results do not reduce by construction to that signal. Overall, the derivation chain is self-contained against external benchmarks.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The central claim depends on the proxy-loss-signal assumption, on the choice of clusters and the initial density threshold, and on the hyperparameters β, α+, and α−, none of which the paper reports precisely.

free parameters (4)
  • Number of clusters K = not reported
    Determines the granularity of per-cluster adaptation. The paper tests baselines with K=10, 100, 500 but does not state which K is used for ADADEDUP or whether it was tuned.
  • Adaptation strength β = not reported numerically
    Controls the magnitude of pruning-ratio adjustment. The paper says it is chosen so the update affects 5-10% of samples, which is a hand-tuned criterion.
  • Scaling factors α+ and α− = not reported
    Required in Algorithm 1 to scale positive versus negative loss differentials; values or selection rule are not given.
  • Initial density threshold τ = not reported
    Determines the initial pruned set D_s0 and the initial per-cluster pruning ratios γ_i; the paper does not specify how τ is set.
assumptions (4)
  • domain assumption The loss differential Δℓ_i = ℓ_i^s − ℓ_i^p computed from a proxy model trained on the kept set is a faithful estimate of the marginal information content of the pruned samples.
    Section 3.2 and Eq. (2). This is the key modeling assumption; if the proxy overfits to kept samples, Δℓ_i is biased.
  • domain assumption K-means clustering on VLM or caption embeddings yields semantic clusters whose internal redundancy is homogeneous.
    Section 3.3.1 Step 1. The cluster-adaptive adjustment only makes sense if clusters are semantically meaningful.
  • domain assumption The proxy model with loss ℓ is a good stand-in for the target model's behavior on pruned samples.
    Section 3.3: 'The proxy model can be the target model intended for final training or a computationally cheaper alternative to guide the pruning process.'
  • standard math The bilevel data-pruning objective (Eq. 1) is NP-hard, so heuristic approximations are acceptable.
    Section 3.1 states the subset selection problem is NP-hard, a standard combinatorial optimization result.

how reviews work

0 comments
Cite this review

Pith. "Pith review of AdaDeDup: Adaptive Hybrid Data Pruning for Efficient Large-Scale Object Detection Training." pith.science (2026). https://pith.science/paper/TLKD7UN6

@misc{pith2026250700049,
  author       = {Pith},
  title        = {Pith review of: AdaDeDup: Adaptive Hybrid Data Pruning for Efficient Large-Scale Object Detection Training},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TLKD7UN6}},
  note         = {Machine review of arXiv:2507.00049}
}
read the original abstract

The computational burden and inherent redundancy of large-scale datasets challenge the training of contemporary machine learning models. Data pruning offers a solution by selecting smaller, informative subsets, yet existing methods struggle: density-based approaches can be task-agnostic, while model-based techniques may introduce redundancy or prove computationally prohibitive. We introduce Adaptive De-Duplication (AdaDeDup), a novel hybrid framework that synergistically integrates density-based pruning with model-informed feedback in a cluster-adaptive manner. AdaDeDup first partitions data and applies an initial density-based pruning. It then employs a proxy model to evaluate the impact of this initial pruning within each cluster by comparing losses on kept versus pruned samples. This task-aware signal adaptively adjusts cluster-specific pruning thresholds, enabling more aggressive pruning in redundant clusters while preserving critical data in informative ones. Extensive experiments on large-scale object detection benchmarks (Waymo, COCO, nuScenes) using standard models (BEVFormer, Faster R-CNN) demonstrate AdaDeDup's advantages. It significantly outperforms prominent baselines, substantially reduces performance degradation (e.g., over 54% versus random sampling on Waymo), and achieves near-original model performance while pruning 20% of data, highlighting its efficacy in enhancing data efficiency for large-scale model training. Code is open-sourced.

Figures

Figures reproduced from arXiv: 2507.00049 by the authors.

Figure 1
Figure 1. Illustration on density-based data pruning [ [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Left: Average distance to its 10 nearest neighbors in visual embeddings for samples in each cluster. A lower distance indicates higher redundancy where more samples will be removed during density-based data pruning. Insights from [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. mAP (± std. dev.) vs. data retained on Waymo. ADADEDUP shows significant perfor￾mance retention [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figures from the paper (2 more)
Figure 5
Figure 5. Figure 5: Performance on nuScenes when pruned to 70% data. Left: Table showing mAP (± std. dev.) and performance drop from full dataset. Right: Visualization of performance drop (mAP ± std. dev.). ADADEDUP demonstrates the smallest performance degradation. 4.4 Results on Pruning…
Figure 6
Figure 6. Figure 6: Illustration on density-based data pruning [ [PITH_FULL_IMAGE:figures/full_fig_p020_6.png]

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Test-Time Coverage: Test-Conditioned Data Curation for Deployment-Aware Learning

    cs.AI 2026-07 conditional novelty 6.0 of 10

    TTCov curates training data for deployment by building an LLM-generated atomic-proposition atlas of the test distribution and greedily selecting clips that match it.

Reference graph

Works this paper leans on

41 extracted references · 15 canonical work pages · cited by 1 Pith paper

  1. [1]

    How much more data do i need? estimating requirements for downstream tasks

    Rafid Mahmood, James Lucas, David Acuna, Daiqing Li, Jonah Philion, Jose M Alvarez, Zhiding Yu, Sanja Fidler, and Marc T Law. How much more data do i need? estimating requirements for downstream tasks. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 275–284, 2022

    work page 2022

  2. [2]

    Semd- edup: Data-efficient learning at web-scale through semantic deduplication.arXiv preprint arXiv:2303.09540, 2023

    Amro Abbas, Kushal Tirumala, Dániel Simig, Surya Ganguli, and Ari S Morcos. Semd- edup: Data-efficient learning at web-scale through semantic deduplication.arXiv preprint arXiv:2303.09540, 2023

    arXiv 2023

  3. [3]

    Optimizing data collection for machine learning.Journal of Machine Learning Research, 26(38):1–52, 2025

    Rafid Mahmood, James Lucas, Jose M Alvarez, Sanja Fidler, and Marc T Law. Optimizing data collection for machine learning.Journal of Machine Learning Research, 26(38):1–52, 2025

    work page 2025

  4. [4]

    Beyond neural scaling laws: beating power law scaling via data pruning.Advances in Neural Information Processing Systems, 35:19523–19536, 2022

    Ben Sorscher, Robert Geirhos, Shashank Shekhar, Surya Ganguli, and Ari Morcos. Beyond neural scaling laws: beating power law scaling via data pruning.Advances in Neural Information Processing Systems, 35:19523–19536, 2022

    2022

  5. [5]

    Sse: Multimodal semantic data selection and enrichment for industrial-scale data assimilation.ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD), 2025

    Maying Shen, Nadine Chang, Sifei Liu, and Jose M Alvarez. Sse: Multimodal semantic data selection and enrichment for industrial-scale data assimilation.ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD), 2025

    work page 2025

  6. [6]

    Effective pruning of web-scale datasets based on complexity of concept clusters

    Amro Abbas, Evgenia Rusak, Kushal Tirumala, Wieland Brendel, Kamalika Chaudhuri, and Ari S Morcos. Effective pruning of web-scale datasets based on complexity of concept clusters. arXiv preprint arXiv:2401.04578, 2024

    arXiv 2024

  7. [7]

    Zero-shot coreset selection: Efficient pruning for unlabeled data.arXiv preprint arXiv:2411.15349, 2024

    Brent A Griffin, Jacob Marks, and Jason J Corso. Zero-shot coreset selection: Efficient pruning for unlabeled data.arXiv preprint arXiv:2411.15349, 2024

    arXiv 2024

  8. [8]

    Efficient coreset selection with cluster-based methods

    Chengliang Chai, Jiayi Wang, Nan Tang, Ye Yuan, Jiabin Liu, Yuhao Deng, and Guoren Wang. Efficient coreset selection with cluster-based methods. InProceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, pages 167–178, 2023

    work page 2023

Show all 41 references
  1. [9]

    Moderate coreset: A universal method of data selection for real-world data-efficient deep learning

    Xiaobo Xia, Jiale Liu, Jun Yu, Xu Shen, Bo Han, and Tongliang Liu. Moderate coreset: A universal method of data selection for real-world data-efficient deep learning. InThe Eleventh International Conference on Learning Representations, 2022

    2022

  2. [10]

    Data pruning via moving-one-sample-out.Advances in neural information processing systems, 36: 18251–18262, 2023

    Haoru Tan, Sitong Wu, Fei Du, Yukang Chen, Zhibin Wang, Fan Wang, and Xiaojuan Qi. Data pruning via moving-one-sample-out.Advances in neural information processing systems, 36: 18251–18262, 2023

    2023

  3. [11]

    Data curation via joint example selection further accelerates multimodal learning.Advances in Neural Information Processing Systems, 37:141240–141260, 2024

    Talfan Evans, Nikhil Parthasarathy, Hamza Merzic, and Olivier Henaff. Data curation via joint example selection further accelerates multimodal learning.Advances in Neural Information Processing Systems, 37:141240–141260, 2024

    2024

  4. [12]

    Selection via proxy: Efficient data selection for deep learning.arXiv preprint arXiv:1906.11829, 2019

    Cody Coleman, Christopher Yeh, Stephen Mussmann, Baharan Mirzasoleiman, Peter Bailis, Percy Liang, Jure Leskovec, and Matei Zaharia. Selection via proxy: Efficient data selection for deep learning.arXiv preprint arXiv:1906.11829, 2019

    1906 arXiv

  5. [13]

    Coreset selection for object detection

    Hojun Lee, Suyoung Kim, Junhoo Lee, Jaeyoung Yoo, and Nojun Kwak. Coreset selection for object detection. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 7682–7691, 2024

    2024

  6. [14]

    Goodcore: Data-effective and data-efficient machine learning through coreset selection over incomplete data.Proceedings of the ACM on Management of Data, 1(2):1–27, 2023

    Chengliang Chai, Jiabin Liu, Nan Tang, Ju Fan, Dongjing Miao, Jiayi Wang, Yuyu Luo, and Guoliang Li. Goodcore: Data-effective and data-efficient machine learning through coreset selection over incomplete data.Proceedings of the ACM on Management of Data, 1(2):1–27, 2023. 10

    2023

  7. [15]

    nuscenes: A multimodal dataset for autonomous driving

    Holger Caesar, Varun Bankiti, Alex H Lang, Sourabh V ora, Venice Erin Liong, Qiang Xu, Anush Krishnan, Yu Pan, Giancarlo Baldan, and Oscar Beijbom. nuscenes: A multimodal dataset for autonomous driving. InProceedings of the IEEE/CVF conference on computer vision and pattern re...

    2020

  8. [16]

    Scalability in perception for autonomous driving: Waymo open dataset

    Pei Sun, Henrik Kretzschmar, Xerxes Dotiwalla, Aurelien Chouard, Vijaysai Patnaik, Paul Tsui, James Guo, Yin Zhou, Yuning Chai, Benjamin Caine, et al. Scalability in perception for autonomous driving: Waymo open dataset. InProceedings of the IEEE/CVF conference on computer vis...

    2020

  9. [17]

    Lvis: A dataset for large vocabulary instance segmentation

    Agrim Gupta, Piotr Dollar, and Ross Girshick. Lvis: A dataset for large vocabulary instance segmentation. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 5356–5364, 2019

    2019

  10. [18]

    Microsoft coco: Common objects in context

    Tsung-Yi Lin, Michael Maire, Serge Belongie, James Hays, Pietro Perona, Deva Ramanan, Piotr Dollár, and C Lawrence Zitnick. Microsoft coco: Common objects in context. InComputer vision–ECCV 2014: 13th European conference, zurich, Switzerland, September 6-12, 2014, proceedings,...

    2014

  11. [19]

    Performance scaling via optimal transport: Enabling data selection from partially revealed sources.Advances in Neural Information Processing Systems, 36:61341–61363, 2023

    Feiyang Kang, Hoang Anh Just, Anit Kumar Sahu, and Ruoxi Jia. Performance scaling via optimal transport: Enabling data selection from partially revealed sources.Advances in Neural Information Processing Systems, 36:61341–61363, 2023

    2023

  12. [20]

    Understanding black-box predictions via influence functions

    Pang Wei Koh and Percy Liang. Understanding black-box predictions via influence functions. InInternational conference on machine learning, pages 1885–1894. PMLR, 2017

    2017

  13. [21]

    Active learning for convolutional neural networks: A core-set approach.arXiv preprint arXiv:1708.00489, 2017

    Ozan Sener and Silvio Savarese. Active learning for convolutional neural networks: A core-set approach.arXiv preprint arXiv:1708.00489, 2017

    2017 arXiv

  14. [22]

    An empirical study of example forgetting during deep neural network learning.arXiv preprint arXiv:1812.05159, 2018

    Mariya Toneva, Alessandro Sordoni, Remi Tachet des Combes, Adam Trischler, Yoshua Bengio, and Geoffrey J Gordon. An empirical study of example forgetting during deep neural network learning.arXiv preprint arXiv:1812.05159, 2018

    2018 arXiv

  15. [23]

    Deep learning on a data diet: Finding important examples early in training.Advances in neural information processing systems, 34:20596–20607, 2021

    Mansheej Paul, Surya Ganguli, and Gintare Karolina Dziugaite. Deep learning on a data diet: Finding important examples early in training.Advances in neural information processing systems, 34:20596–20607, 2021

    2021

  16. [24]

    Coresets via bilevel optimization for continual learning and streaming.Advances in neural information processing systems, 33: 14879–14890, 2020

    Zalán Borsos, Mojmir Mutny, and Andreas Krause. Coresets via bilevel optimization for continual learning and streaming.Advances in neural information processing systems, 33: 14879–14890, 2020

    2020

  17. [25]

    Gradient-based bi-level optimization for deep learning: A survey.arXiv preprint arXiv:2207.11719, 2022

    Can Chen, Xi Chen, Chen Ma, Zixuan Liu, and Xue Liu. Gradient-based bi-level optimization for deep learning: A survey.arXiv preprint arXiv:2207.11719, 2022

    2022 arXiv

  18. [26]

    Springer Science & Business Media, 1998

    Jonathan F Bard.Practical bilevel optimization: algorithms and applications, volume 30. Springer Science & Business Media, 1998

    1998

  19. [27]

    Datamodels: Predicting predictions from training data.arXiv preprint arXiv:2202.00622, 2022

    Andrew Ilyas, Sung Min Park, Logan Engstrom, Guillaume Leclerc, and Aleksander Madry. Datamodels: Predicting predictions from training data.arXiv preprint arXiv:2202.00622, 2022

    2022 arXiv

  20. [28]

    Autoscale: Automatic prediction of compute-optimal data composition for training llms.arXiv preprint arXiv:2407.20177, 2024

    Feiyang Kang, Yifan Sun, Bingbing Wen, Si Chen, Dawn Song, Rafid Mahmood, and Ruoxi Jia. Autoscale: Automatic prediction of compute-optimal data composition for training llms.arXiv preprint arXiv:2407.20177, 2024

    2024

  21. [29]

    Bevformer: learning bird’s-eye-view representation from lidar-camera via spatiotemporal transformers.IEEE Transactions on Pattern Analysis and Machine Intelligence, 2024

    Zhiqi Li, Wenhai Wang, Hongyang Li, Enze Xie, Chonghao Sima, Tong Lu, Qiao Yu, and Jifeng Dai. Bevformer: learning bird’s-eye-view representation from lidar-camera via spatiotemporal transformers.IEEE Transactions on Pattern Analysis and Machine Intelligence, 2024

    2024

  22. [30]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. InProceedings of the IEEE conference on computer vision and pattern recognition, pages 770–778, 2016

    2016

  23. [31]

    Faster r-cnn: Towards real-time object detection with region proposal networks.IEEE transactions on pattern analysis and machine intelligence, 39(6):1137–1149, 2016

    Shaoqing Ren, Kaiming He, Ross Girshick, and Jian Sun. Faster r-cnn: Towards real-time object detection with region proposal networks.IEEE transactions on pattern analysis and machine intelligence, 39(6):1137–1149, 2016. 11

    2016

  24. [32]

    Detectron2

    Yuxin Wu, Alexander Kirillov, Francisco Massa, Wan-Yen Lo, and Ross Girshick. Detectron2. https://github.com/facebookresearch/detectron2, 2019

    2019

  25. [33]

    Learning transferable visual models from natural language supervision

    Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, et al. Learning transferable visual models from natural language supervision. InInternational conference on machine learning, pag...

    2021

  26. [34]

    Grounding dino: Marrying dino with grounded pre-training for open-set object detection

    Shilong Liu, Zhaoyang Zeng, Tianhe Ren, Feng Li, Hao Zhang, Jie Yang, Qing Jiang, Chunyuan Li, Jianwei Yang, Hang Su, et al. Grounding dino: Marrying dino with grounded pre-training for open-set object detection. InEuropean Conference on Computer Vision, pages 38–55. Springer, 2024

    2024

  27. [35]

    Improved baselines with visual instruction tuning

    Haotian Liu, Chunyuan Li, Yuheng Li, and Yong Jae Lee. Improved baselines with visual instruction tuning. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 26296–26306, 2024

    2024

  28. [36]

    Data distillation: A survey.arXiv preprint arXiv:2301.04272, 2023

    Noveen Sachdeva and Julian McAuley. Data distillation: A survey.arXiv preprint arXiv:2301.04272, 2023

    2023 arXiv

  29. [37]

    Glister: Generalization based data subset selection for efficient and robust learning

    Krishnateja Killamsetty, Durga Sivasubramanian, Ganesh Ramakrishnan, and Rishabh Iyer. Glister: Generalization based data subset selection for efficient and robust learning. InPro- ceedings of the AAAI Conference on Artificial Intelligence, volume 35, pages 8110–8118, 2021

    2021

  30. [38]

    Grad-match: Gradient matching based data subset selection for efficient deep model training

    Krishnateja Killamsetty, Sivasubramanian Durga, Ganesh Ramakrishnan, Abir De, and Rishabh Iyer. Grad-match: Gradient matching based data subset selection for efficient deep model training. InInternational Conference on Machine Learning, pages 5464–5474. PMLR, 2021

    2021

  31. [39]

    Lava: Data valuation without pre-specified learning algorithms.arXiv preprint arXiv:2305.00054, 2023

    Hoang Anh Just, Feiyang Kang, Jiachen T Wang, Yi Zeng, Myeongseob Ko, Ming Jin, and Ruoxi Jia. Lava: Data valuation without pre-specified learning algorithms.arXiv preprint arXiv:2305.00054, 2023

    2023 arXiv

  32. [40]

    Estimating training data influence by tracing gradient descent.Advances in Neural Information Processing Systems, 33: 19920–19930, 2020

    Garima Pruthi, Frederick Liu, Satyen Kale, and Mukund Sundararajan. Estimating training data influence by tracing gradient descent.Advances in Neural Information Processing Systems, 33: 19920–19930, 2020

    2020

  33. [41]

    informative

    Feiyang Kang, Hoang Anh Just, Yifan Sun, Himanshu Jahagirdar, Yuanzhi Zhang, Rongxing Du, Anit Kumar Sahu, and Ruoxi Jia. Get more for less: Principled data selection for warming up fine-tuning in llms.arXiv preprint arXiv:2405.02774, 2024. 12 Appendices A Extended Related Wor...

    2024 arXiv

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.