Pith. sign in

REVIEW 5 major objections 4 minor 28 references

Act, Think or Abstain: Complexity-Aware Adaptive Inference for Vision-Language-Action Models

T0 review · 5 major / 4 minor · reviewed 2026-08-02 · deepseek-v4-flash

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

Pith's one-line read A vision-language-action model can read the difficulty of a task from its own visual embeddings and decide on the fly whether to act immediately, reason first, or refuse to act.

desk verdict Useful adaptive-routing idea with a vision-only finding that matters, but the OOD evaluation is in-sample and the write-up contradicts itself. read the letter →

arxiv 2603.05147 v2 pith:SYJVEKJ2 submitted 2026-03-05 cs.CV cs.RO

Riccardo Andrea Izzo , Gianluca Bardaro , Matteo Matteucci This is my paper · ORCID
classification cs.CVcs.RO
keywords adaptiveinferencevision-language-actionmodelsout-of-distributiondetectionGaussianmixturetaskcomplexitysaferobotexecutionmixupembeddingscoring
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 sets out to show that a robot policy need not spend the same amount of computation on every instruction. It describes a lightweight gating module that scores how novel a perceived state is using the latent embeddings that a vision-language-action model already produces, then routes the episode into one of three paths: execute immediately (Act), perform one round of additional reasoning (Think), or stop before acting (Abstain). The authors find that visual embeddings alone, scored by a Gaussian mixture model, give the most reliable complexity signal, reaching a macro F1-score of 84.34% on separating the three classes and retaining strong performance with only 5% of training data. If the approach holds, it offers a cheap, built-in safeguard for embodied foundation models and a way to budget reasoning only when it is actually needed.

What carries the argument

The core mechanism is a routing module built on the VLA's pretrained vision-language backbone. Visual and text embeddings are spatially averaged, projected with PCA to 64 dimensions, and then scored by two density estimators: a Gaussian mixture model with three components, scored by the covariance-aware distance to the closest component, and a 1-nearest-neighbour Euclidean-distance scorer. The resulting four scores are concatenated into a vector and passed through a small multilayer perceptron that outputs a probability over three strategies: Act, Think, or Abstain. Partially out-of-distribution training examples for the Think class are synthesized by mixup interpolation between in-distribut

What would settle it

Collect a held-out set of partial out-of-distribution tasks not generated by mixup, such as unseen object colors, positions, or lighting changes that alter the task meaning but leave it recoverable, and measure the router's Think-path recall on them. If recall falls near the baseline's 14% instead of the paper's claimed F1 range, then the mixup interpolation does not span the real ambiguity manifold and the central claim collapses.

Watch

Extended reading notes

Core claim

The paper claims that task complexity for a vision-language-action model can be inferred directly from its own vision backbone's latent embeddings, without an external reward model or separate failure detector. A Gaussian mixture model fitted to averaged visual patch embeddings, scored by the Mahalanobis distance to the nearest mixture component, classifies states into in-distribution, partially out-of-distribution, and fully out-of-distribution with 84.34% macro F1 over a balanced validation set. In simulation, this vision-only router selects the Abstain path in 29 of 30 fully out-of-distribution trials, cutting average time spent on failed episodes from roughly 150 seconds to about 3 secon

Load-bearing premise

The whole routing rests on assuming that interpolating between known and unknown feature vectors with Eq. (9) produces features that faithfully stand in for genuinely ambiguous real-world states; if those synthetic features are artifacts, the Think path is trained on the wrong target and the reported safety gains may not transfer to unseen distribution shifts.

Editorial extensions

If this is right

  • A router can be added on top of an existing VLA without retraining the action policy, using only embeddings the backbone already computes.
  • Easy tasks keep low latency, ambiguous tasks get one extra reasoning pass, and impossible tasks are stopped early, so average inference time can drop sharply instead of rising uniformly.
  • Because vision-only scoring is best, the language branch can be ignored during gating, simplifying the detector and reducing its computational cost.
  • The detector reaches near-peak performance with roughly 5% of the training data, suggesting that adapting to a new robot or environment needs only a small calibration set.

Reading between the lines

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

  • The success of vision-only complexity scoring suggests the same gating idea could be applied to any visual policy, not just VLAs, by attaching a density estimator to a frozen image encoder.
  • The three-way hard split is likely the weakest point: real partial shifts blend continuously into full shifts, so a continuous risk score or calibrated probability would probably route more safely at the boundaries.
  • The mixup-generated Think features are the main unvalidated piece; a test on naturally occurring partial out-of-distribution states would tell whether the synthetic interpolation captures real ambiguity or only artifacts of linear mixing.
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

5 major / 4 minor

Summary. The paper proposes an adaptive inference framework for Vision-Language-Action (VLA) models. It uses embeddings from the VLM backbone of SmolVLA, scores them with a Gaussian Mixture Model (GMM) and a k-nearest-neighbour (kNN) estimator, and feeds the scores to a small MLP that routes each episode into one of three strategies: Act (in-distribution execution), Think (partial out-of-distribution, triggers extra reasoning), or Abstain (full out-of-distribution, halts execution). Training uses LIBERO as in-distribution, LIBERO-PRO as partial OOD, and NYU/CMU Franka datasets as full OOD, with a mixup strategy (Eq. 9) to synthesize intermediate 'Think' features. Experiments on LIBERO/LIBERO-PRO and a SO-ARM 101 robot report high macro F1 for the vision-only GMM configuration (84.34%, Section IV-C) and large numbers of prevented failures on swap/task variants (Table I).

Significance. If the central claims hold, the framework would be a practical contribution: it offers a lightweight, model-agnostic way to gate VLA inference by task complexity, with density-based uncertainty scores rather than full reasoning at every step. The paper provides useful ablations over modality and data fraction, and reports a real-robot deployment, which strengthens the practical case. The release of code and models is also a positive. However, the current manuscript contains internal contradictions between the abstract and the body, and the safety evaluation is contaminated by training/evaluation overlap. As written, the central evidence does not yet support the claimed generalization to unseen out-of-distribution scenarios.

major comments (5)
  1. [Abstract vs. Section IV-C] The abstract states that a fused vision-language configuration achieves up to 87.5% F1 across SmolVLA and π0, while the full text's abstract and Section IV-C report that vision-only GMM is best (84.34% macro F1) and that text/fused features are detrimental. No π0 experiments appear anywhere in the paper. This is a direct internal contradiction on the paper's main quantitative result and must be resolved. The reported F1 values, the backbone list, and the abstract's 'surpasses state-of-the-art failure detectors' claim (which is never evaluated in the experiments) need to be aligned with the actual experiments and removed if unsupported.
  2. [Section III-D and Table I] The OOD evaluation is circular. Training defines LIBERO as ID, LIBERO-PRO as partially OOD (Think), and NYU/CMU Franka as fully OOD (Abstain). Section IV-D then evaluates swap and task variants of LIBERO-PRO as OOD and reports 28-30/30 Abstain decisions in Table I. These swap/task variants are part of the same LIBERO-PRO dataset used to train the Think class, so the near-total Abstain rate does not demonstrate generalization to unseen distribution shifts. The claim that the framework 'preemptively halts execution when encountering significant physical or semantic anomalies' requires evaluation on a genuinely held-out OOD set, not on training-corpus variants with contradictory labels.
  3. [Section III-D, Eq. (9)] The mixup procedure for the Think class is underspecified. Equation (9) interpolates in raw feature space (z_think = λ z_ID + (1-λ) z_OOD), but the MLP classifier is trained on the score vector u of Eq. (6), which contains GMM and kNN scores. The manuscript does not explain how the synthetic z_think features are converted into scores, whether the same PCA/GMM/kNN pipeline is applied to the mixed features, or whether the mixup is instead applied to the score vector. Since the Think class is entirely built on this synthetic interpolation, the exact training procedure is load-bearing and must be described concretely.
  4. [Conclusions vs. Table I] The conclusion claims that 'our framework never mistakenly abstained from executing in-distribution tasks' and that the Think path 'effectively increases performance.' Both statements are contradicted by the authors' own Table I. For example, in the base (in-distribution) rows, the Goal suite shows one Abstain out of 30, Spatial shows five Abstains, and Long shows ten. In partially OOD variants, the number of Think decisions is often only 1-3 per 30 episodes, and success rates are nearly unchanged relative to the SmolVLA baseline (e.g., Goal-language: 90.00 vs 90.00, Object-language: 90.67 vs 96.67). The claims should be reworded to match the reported numbers, and the success-rate benefits of Think need a proper controlled comparison.
  5. [Section IV-D and Abstract] The abstract claims the method 'surpasses state-of-the-art failure detectors,' but no comparison with any existing failure detector (e.g., SAFE or similar) is reported in the experimental section. The paper also does not report the 87.5% F1 or π0 results that appear in the abstract. These unsupported claims must be either backed with experiments or removed.
minor comments (4)
  1. [Throughout] There are several typographical issues: 'out-of-distibution' in Figure 1, '00.0±0.0' in Table I, inconsistent use of 'k' vs 'K' for the number of GMM components, and 'π 0' formatting. A careful proofread would improve readability.
  2. [Section III-C] The score vector u in Eq. (6) is defined as a 4-dimensional vector but only four scores are listed (three GMM scores and one kNN). The text says 'visual, text, and fused representations' but the kNN score is explicitly visual-only. This is fine, but it should be clarified earlier that kNN is not used for text or fused features.
  3. [Section IV-B] The data-scaling plot (Figure 3) is described in the text but the exact number of training samples at each fraction is only stated for the 1% case ('less than 1,000 samples'). Reporting exact sample counts in the caption or text would make the scaling behavior more interpretable.
  4. [Section VI] The limitations are candid and useful, but the statement that 'this assumption relies on having the corresponding LIBERO distribution as the ID reference and does not generalize across VLA models' actually underscores the circularity concern in Experiment D. It might be better to move this caveat earlier, near Section III-D, and to acknowledge that the current OOD results are in-sample for the detector.

Circularity Check

0 steps flagged · score 2.0 of 10

No load-bearing circularity; the classifier is trained and validated on explicit splits, with the main caveat being an ambiguous LIBERO-PRO Think/Abstain label assignment in the simulation evaluation.

full rationale

The paper does not claim a first-principles derivation; it trains a supervised router. The training labels are explicit: ID = LIBERO, partially-OOD = LIBERO-PRO (with a mixup-generated Think prototype via Eq. 9), and fully-OOD = NYU/CMU Franka. The GMM/kNN scores are standard density estimates and the MLP is trained with a 50/25/25 partition, so the reported Macro F1 scores are held-out evaluations rather than re-predictions of the training labels. There are no load-bearing self-citations, no imported uniqueness theorems, and no fitted parameter that is renamed as a prediction. The mixup equation defines the Think prototype, but the test signal is evaluated on LIBERO, LIBERO-PRO, and real-robot tasks, so it does not reduce to the same training inputs. The most serious concern is an evaluation-label inconsistency in Sec. IV-D: LIBERO-PRO is described as the Think distribution and then the same benchmark's failing variants are used to measure the Abstain policy's prevented failures (PF in Table I), making the OOD safety test partly in-family rather than truly held-out. That is a correctness/validity concern, not a circular reduction by construction, because the classifier's decisions are not forced by its training labels and the paper also provides held-out validation and real-robot OOD evidence.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new physical entities. Its main dependencies are standard OOD assumptions plus three domain-specific proxies: LIBERO-PRO as partial OOD, other Franka datasets as full OOD, and mixup as a generator of intermediate states. The mixup assumption is the most fragile.

free parameters (5)
  • Number of GMM components K = 3
    Chosen via validation ablation; performance peaks at K=3 (Figure 2). This is a model selection parameter fitted to validation data.
  • PCA target dimension D' = 64
    Chosen to preserve 95% of variance; exact dimension is a hyperparameter.
  • Shrinkage coefficient rho = 0.01
    Default in Ledoit-Wolf covariance shrinkage, not tuned (Eq. 2).
  • Mixup Beta shape parameters = Beta(0.5, 0.5)
    Used to synthesize the Think class (Eq. 9); no sensitivity analysis provided.
  • MLP hidden sizes = 64 and 32
    Architecture chosen without reported comparison.
assumptions (4)
  • domain assumption VLM backbone embeddings contain sufficient information about task complexity and OOD-ness.
    Whole method depends on distances in embedding space reflecting task difficulty; no independent validation of this mapping beyond the paper's own classifier.
  • domain assumption LIBERO-PRO defines a valid partially-OOD distribution for any VLA fine-tuned on LIBERO.
    Used to train and evaluate the Think path; the paper notes this assumption 'does not generalize across VLA models' (Section III-D).
  • ad hoc to paper Mixup interpolation between ID and OOD embedding features generates realistic partially-OOD states.
    Eq. 9 creates the Think training signal; there is no evidence that linear interpolations in latent space correspond to genuinely ambiguous scenes.
  • domain assumption The base VLA policy is fixed during training of the router.
    The router is trained without updating the policy, so the 'Think' improvements rely on the pre-trained policy's ability to use the added scene cues.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Act, Think or Abstain: Complexity-Aware Adaptive Inference for Vision-Language-Action Models." pith.science (2026). https://pith.science/paper/SYJVEKJ2

@misc{pith2026260305147,
  author       = {Pith},
  title        = {Pith review of: Act, Think or Abstain: Complexity-Aware Adaptive Inference for Vision-Language-Action Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SYJVEKJ2}},
  note         = {Machine review of arXiv:2603.05147}
}
abstract

Current research on Vision-Language-Action (VLA) models predominantly focuses on enhancing generalization through reasoning techniques. While effective, these improvements increase computational complexity and inference latency. Furthermore, these mechanisms are typically applied indiscriminately, wasting resources on trivial tasks while failing to provide the uncertainty estimation necessary to prevent catastrophic failure on out-of-distribution scenarios. Inspired by human cognition, we propose an adaptive framework that dynamically routes VLA execution based on the complexity of the perceived state. Our approach transforms the VLA's vision-language backbone into an active detection tool by projecting latent embeddings into a set of parametric and non-parametric estimators. This allows the system to execute known tasks immediately (Act), reason about ambiguous scenarios (Think), and preemptively halt execution when encountering physical or semantic anomalies (Abstain). We find that a Gaussian Mixture Model fitted to fused vision-language embeddings provides the most reliable task-complexity signal, combining visual novelty with instruction context and cross-modal compatibility. Evaluated on the LIBERO and LIBERO-PRO benchmarks as well as on a real robot, our fused configuration achieves up to 87.5% F1-score across two VLA backbones (SmolVLA and $\pi_0$), retains 83% with as little as 5% of training data, and surpasses state-of-the-art failure detectors.

Figures

Figures reproduced from arXiv: 2603.05147 by the authors.

Figure 1
Figure 1. Overview of our framework. Given an image observation and a language instruction of a task, the system extracts text and vision embeddings from the SmolVLA backbone. These embeddings are scored using an ensemble of GMMs and kNN. The resulting scores are consolidated into a unified score vector and processed by a custom MLP to select the optimal execution strategy. High confidence triggers direct execution (Act), whi… view at source ↗
Figure 2
Figure 2. Impact of GMM components k on OOD detection. The F1-score peaks at k = 3, demonstrating that while multiple Gaussians are necessary to capture the task manifold’s complexity, further increases lead to overfitting and diminishing returns. Variance remains stable across the tested range. • RQ1: What are the individual contributions of GMM, kNN, and the MLP across different modalities in de￾tecting task distribution sh… view at source ↗
Figure 3
Figure 3. Data Scaling and Component Ablation. Macro F1-Score performance across increasing fractions of training data for different model architectures under ID (LIBERO), partially OOD (LIBERO-PRO), and fully OOD tasks [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Confusion matrices across configurations. The GMM (vision-only) achieves the best class separation with zero leakage from fully OOD tasks into the "Act" regime. Conversely, the Baseline MLP and multimodal variants show high "Act" bias when evaluating partially OOD task…
Figure 5
Figure 5. Figure 5: For the evaluation metrics, we considered Success [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 5
Figure 5. Figure 5: Rollout examples on simulation. Representative episodes from the LIBERO and LIBERO-PRO benchmarks illustrating our adaptive framework. Known tasks are executed with high confidence (left), ambiguous scenarios trigger additional reasoning leading to success later on (ce…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

28 extracted references · 9 linked inside Pith

  1. [1]

    Robotic control via embodied chain-of-thought reasoning,

    M. Zawalski, W. Chen, K. Pertsch, O. Mees, C. Finn, and S. Levine, “Robotic control via embodied chain-of-thought reasoning,” inCon- ference on Robot Learning. PMLR, 2025, pp. 3157–3181

    2025

  2. [2]

    Cot-vla: Visual chain-of-thought reasoning for vision-language-action models,

    Q. Zhao, Y . Lu, M. J. Kim, Z. Fu, Z. Zhang, Y . Wu, Z. Li, Q. Ma, S. Han, C. Finnet al., “Cot-vla: Visual chain-of-thought reasoning for vision-language-action models,” inProceedings of the Computer Vision and Pattern Recognition Conference, 2025, pp. 1702–1713

    2025

  3. [3]

    Fast ecot: Efficient embodied chain-of-thought via thoughts reuse,

    Z. Duan, Y . Zhang, S. Geng, G. Liu, J. Boedecker, and C. X. Lu, “Fast ecot: Efficient embodied chain-of-thought via thoughts reuse,”arXiv preprint arXiv:2506.07639, 2025

    arXiv 2025

  4. [4]

    What matters in building vision–language– action models for generalist robots,

    X. Li, P. Li, L. Qian, M. Liu, D. Wang, J. Liu, B. Kang, X. Ma, X. Wang, D. Guoet al., “What matters in building vision–language– action models for generalist robots,”Nature Machine Intelligence, pp. 1–15, 2026

    2026

  5. [5]

    Bridgedata v2: A dataset for robot learning at scale,

    H. R. Walke, K. Black, T. Z. Zhao, Q. Vuong, C. Zheng, P. Hansen- Estruch, A. W. He, V . Myers, M. J. Kim, M. Duet al., “Bridgedata v2: A dataset for robot learning at scale,” inConference on Robot Learning. PMLR, 2023, pp. 1723–1736

    2023

  6. [6]

    Open x-embodiment: Robotic learning datasets and rt-x models: Open x- embodiment collaboration 0,

    A. O’Neill, A. Rehman, A. Maddukuri, A. Gupta, A. Padalkar, A. Lee, A. Pooley, A. Gupta, A. Mandlekar, A. Jainet al., “Open x-embodiment: Robotic learning datasets and rt-x models: Open x- embodiment collaboration 0,” in2024 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2024, pp. 6892–6903

    2024

  7. [7]

    Rt-2: Vision-language-action models transfer web knowledge to robotic control,

    B. Zitkovich, T. Yu, S. Xu, P. Xu, T. Xiao, F. Xia, J. Wu, P. Wohlhart, S. Welker, A. Wahidet al., “Rt-2: Vision-language-action models transfer web knowledge to robotic control,” inConference on Robot Learning. PMLR, 2023, pp. 2165–2183

    2023

  8. [8]

    Open- vla: An open-source vision-language-action model,

    M. J. Kim, K. Pertsch, S. Karamcheti, T. Xiao, A. Balakrishna, S. Nair, R. Rafailov, E. Foster, G. Lam, P. Sanketiet al., “Open- vla: An open-source vision-language-action model,”arXiv preprint arXiv:2406.09246, 2024

    arXiv 2024

Show all 28 references
  1. [9]

    pi_0: A vision- language-action flow model for general robot control,

    K. Black, N. Brown, D. Driess, A. Esmail, M. Equi, C. Finn, N. Fusai, L. Groom, K. Hausman, B. Ichteret al., “pi_0: A vision- language-action flow model for general robot control,”arXiv preprint arXiv:2410.24164, 2024

    2024 arXiv

  2. [10]

    Gr00t n1: An open foundation model for generalist humanoid robots,

    J. Bjorck, F. Castañeda, N. Cherniadev, X. Da, R. Ding, L. Fan, Y . Fang, D. Fox, F. Hu, S. Huanget al., “Gr00t n1: An open foundation model for generalist humanoid robots,”arXiv preprint arXiv:2503.14734, 2025

    2025 arXiv

  3. [11]

    π ∗ 0.6: a vla that learns from experience,

    P. Intelligence, A. Amin, R. Aniceto, A. Balakrishna, K. Black, K. Conley, G. Connors, J. Darpinian, K. Dhabalia, J. DiCarloet al., “π ∗ 0.6: a vla that learns from experience,” 2025

    2025

  4. [12]

    Vla-rl: Towards masterful and general robotic ma- nipulation with scalable reinforcement learning,

    G. Lu, W. Guo, C. Zhang, Y . Zhou, H. Jiang, Z. Gao, Y . Tang, and Z. Wang, “Vla-rl: Towards masterful and general robotic ma- nipulation with scalable reinforcement learning,”arXiv preprint arXiv:2505.18719, 2025

    2025 arXiv

  5. [13]

    Smolvla: A vision-language-action model for affordable and efficient robotics,

    M. Shukor, D. Aubakirova, F. Capuano, P. Kooijmans, S. Palma, A. Zouitine, M. Aractingi, C. Pascal, M. Russi, A. Marafioti, S. Alibert, M. Cord, T. Wolf, and R. Cadene, “Smolvla: A vision-language-action model for affordable and efficient robotics,”

  6. [14]

    Tinyvla: Towards fast, data-efficient vision- language-action models for robotic manipulation,

    J. Wen, Y . Zhu, J. Li, M. Zhu, Z. Tang, K. Wu, Z. Xu, N. Liu, R. Cheng, C. Shenet al., “Tinyvla: Towards fast, data-efficient vision- language-action models for robotic manipulation,”IEEE Robotics and Automation Letters, 2025

    2025

  7. [15]

    Instructvla: Vision-language-action instruction tuning from understanding to manipulation,

    S. Yang, H. Li, Y . Chen, B. Wang, Y . Tian, T. Wang, H. Wang, F. Zhao, Y . Liao, and J. Pang, “Instructvla: Vision-language-action instruction tuning from understanding to manipulation,”arXiv preprint arXiv:2507.17520, 2025

    2025

  8. [16]

    Steerable vision-language- action policies for embodied reasoning and hierarchical control,

    W. Chen, J. S. Bhatia, C. Glossop, N. Mathihalli, R. Doshi, A. Tang, D. Driess, K. Pertsch, and S. Levine, “Steerable vision-language- action policies for embodied reasoning and hierarchical control,”arXiv preprint arXiv:2602.13193, 2026

    2026 arXiv

  9. [17]

    Onetwovla: A unified vision-language-action model with adaptive reasoning,

    F. Lin, R. Nai, Y . Hu, J. You, J. Zhao, and Y . Gao, “Onetwovla: A unified vision-language-action model with adaptive reasoning,”arXiv preprint arXiv:2505.11917, 2025

    2025

  10. [18]

    Safe: Multitask failure detection for vision-language- action models,

    Q. Gu, Y . Ju, S. Sun, I. Gilitschenski, H. Nishimura, M. Itkina, and F. Shkurti, “Safe: Multitask failure detection for vision-language- action models,” inThe Thirty-ninth Annual Conference on Neural Information Processing Systems, 2025

    2025

  11. [19]

    Smolvlm: Redefining small and efficient multimodal models,

    A. Marafioti, O. Zohar, M. Farré, M. Noyan, E. Bakouch, P. Cuenca, C. Zakka, L. B. Allal, A. Lozhkov, N. Taziet al., “Smolvlm: Redefining small and efficient multimodal models,”arXiv preprint arXiv:2504.05299, 2025

    2025 arXiv

  12. [20]

    Flow matching for generative modeling,

    Y . Lipman, R. T. Chen, H. Ben-Hamu, M. Nickel, and M. Le, “Flow matching for generative modeling,” inThe Eleventh International Conference on Learning Representations, 2023

    2023

  13. [21]

    The llama 3 herd of models,

    A. Grattafiori, A. Dubey, A. Jauhri, A. Pandey, A. Kadian, A. Al- Dahle, A. Letman, A. Mathur, A. Schelten, A. Vaughanet al., “The llama 3 herd of models,”arXiv preprint arXiv:2407.21783, 2024

    2024 arXiv

  14. [22]

    Gaussian mixture models

    D. A. Reynoldset al., “Gaussian mixture models.”Encyclopedia of biometrics, vol. 741, no. 659-663, p. 3, 2009

    2009

  15. [23]

    The ma- halanobis distance,

    R. De Maesschalck, D. Jouan-Rimbaud, and D. L. Massart, “The ma- halanobis distance,”Chemometrics and intelligent laboratory systems, vol. 50, no. 1, pp. 1–18, 2000

    2000

  16. [24]

    A well-conditioned estimator for large- dimensional covariance matrices,

    O. Ledoit and M. Wolf, “A well-conditioned estimator for large- dimensional covariance matrices,”Journal of multivariate analysis, vol. 88, no. 2, pp. 365–411, 2004

    2004

  17. [25]

    Libero: Benchmarking knowledge transfer for lifelong robot learn- ing,

    B. Liu, Y . Zhu, C. Gao, Y . Feng, Q. Liu, Y . Zhu, and P. Stone, “Libero: Benchmarking knowledge transfer for lifelong robot learn- ing,”Advances in Neural Information Processing Systems, vol. 36, pp. 44 776–44 791, 2023

    2023

  18. [26]

    Libero-pro: Towards robust and fair evaluation of vision-language-action models beyond memorization,

    X. Zhou, Y . Xu, G. Tie, Y . Chen, G. Zhang, D. Chu, P. Zhou, and L. Sun, “Libero-pro: Towards robust and fair evaluation of vision-language-action models beyond memorization,”arXiv preprint arXiv:2510.03827, 2025

    2025 arXiv

  19. [27]

    mixup: Beyond empirical risk minimization,

    H. Zhang, M. Cisse, Y . N. Dauphin, and D. Lopez-Paz, “mixup: Beyond empirical risk minimization,” inInternational Conference on Learning Representations, 2018

    2018

  20. [2025]

    Available: https://arxiv.org/abs/2506.01844

    [Online]. Available: https://arxiv.org/abs/2506.01844

    arXiv

Pith tools

Reviewed August 2, 2026 · model on record in the stance chip above.