Pith.

Every reviewed paper

Each paper is listed with what Pith's machine review of it concluded.

499,167

papers reviewed

live · read 16 Nov 2026, 06:29 UTC

Narrow the list

Page 26. Ordered by the date each paper was submitted, newest first.

  1. The paper:

    RMIT-ADM+S at TREC RAG 2026: Research Agents, Retriever Ablations, and Open-Weight Replication

    Computer Science · arXiv:2611.16425

    Oleg Zendel, Kun Ran, Chenglong Ma and 2 more

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up. A solid TREC system report with transparent ablations and honest caveats; the open‑weight preference claim is real but the paper could have checked its robustness to a single dominant topic.

    2 major objections, the first at 5.2 / 5.3 · 4 minor

  2. The paper:

    PhysCapBench: A Cross-Domain Benchmark for Physically Grounded Video Captioning

    Computer Science · arXiv:2611.16424

    Haofeng Wang, Pengcheng Wen, Ruiqi Li and 12 more

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    The passes disagreed. Reader: holds up. Skeptic: holds up only with conditions. Referee: major revision.

    A solid benchmark for physical caption quality across three video domains; the main weakness (single LLM judge) is acknowledged and partially mitigated by downstream experiments.

    2 major objections, the first at Section 5.1, Section 6.2 · 6 minor

  3. The paper:

    VOKE: Visual Object Kinematics Estimator for Dexterous In-Hand Manipulation

    Computer Science · arXiv:2611.16423

    Shengjia Zhu, Wenqiang Xu, Lin Shao

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · of version 1 of the paper; version 2 is newer · what was read was not recorded

    Holds up, with conditions. VOKE is a well-engineered, modular framework for estimating object velocity and acceleration from video, with convincing simulation results and real-world transfer; the main caveat is that its real-dataset evaluations rely on spline-derived kinematics references whose accuracy is not independently validated.

    3 major objections, the first at Section IV-A, Paragraph 4 (Motion-quantity reference … · 4 minor

  4. The paper:

    BlinkMimic: Training a Humanoid Motion Tracker in Minutes on a Single GPU

    Computer Science · arXiv:2611.16422

    Fuyi Chen

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up. BlinkMimic delivers a real speedup for humanoid motion tracking with careful engineering and honest limitations; the main open question is whether the world-model gradient biases matter more than the evidence shows.

    2 major objections, the first at Section 5.4, Table 2, Figure 5 · 4 minor

  5. The paper:

    EP250226a: A fast X-ray transient associated with GRB 250226A and its multiwavelength afterglow properties

    Physics · arXiv:2611.16421

    Rui-Zhi Li, Jirong Mao, Aishwarya Linesh Thakur and 64 more

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    The passes disagreed. Reader: holds up. Skeptic: agrees with the reader. Referee: major revision.

    Solid multiwavelength study of a new EP GRB, but the paper's claim of achromatic decay is not formally tested and the data show a ~3.6σ difference in optical vs. X-ray decay indices that should be discussed.

    3 major objections, the first at §3.2 · 5 minor

  6. The paper:

    I See What You Want: Decoding Hierarchical Operator Intent from Visual Teleoperation History

    Computer Science · arXiv:2611.16419

    Xiangyi Jia, Kaixian Qu, Shengzhi Wang and 3 more

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    The passes disagreed. Reader: holds up. Skeptic: agrees with the reader. Referee: major revision.

    Interesting new problem and dataset for hierarchical intent prediction, but the central evaluation rests on annotations generated by GPT-5.6 Sol with only ~50% human review—enough to call the main numbers into question until the authors provide quality controls.

    3 major objections, the first at Sec. IV-A: Annotation pipeline · 7 minor

  7. The paper:

    Model Distribution-Aware Multimodal Dataset Distillation

    Computer Science · arXiv:2611.16418

    Jiajun Shen, Yonglin Wu, Ruonan Yu and 1 more

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up. Solid empirical paper on multimodal dataset distillation; the dynamic queue is the real contribution, and the results hold up.

    3 major objections, the first at §4.4, Figure 4 and §F.1 · 7 minor

  8. The paper:

    VBot Exo2Ego Policy: Lightweight Data Collection and Exocentric-to-Egocentric Video Generation

    Computer Science · arXiv:2611.16417

    Tong Rao, Zhihao Zhang, Yuntian Wang and 3 more

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up, with conditions. Solid engineering with credible gains over EgoX, but the camera conditioning claim lacks a controlled ablation on the main evaluation set—the only evidence is descriptive.

    3 major objections, the first at Section 6.4, Table 7, Appendix A.2 · 5 minor

  9. The paper:

    Synergizing Data-Driven Learning and Physical Rendering for Radio Map Prediction

    Systems · arXiv:2611.16416

    Junyu Jiang, Jing Wang, Yantao Wang and 4 more

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up, with conditions. Novel architecture (SRP + closed-form renderer + flow matching) shows clear gains on simulated data, but the renderer's physical grounding is weaker than claimed and the load-bearing assumption remains unvalidated outside the training distribution.

    3 major objections, the first at Abstract and Section III.C (Eqs. 5–8) · 6 minor

  10. The paper:

    Development and Experimental Characterization of a Compact 228Th–D2O Photoneutron Source for Low-Energy Nuclear Recoil Calibration

    Physics · arXiv:2611.16415

    Yueqiang Tian, Yuxin Cui, Xunan Guo and 5 more

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up, with conditions. Solid engineering paper with a central result that depends on unvalidated simulated efficiency; worth publishing after a clearer discussion of the absolute calibration caveat.

    3 major objections, the first at Section 3, Eq. (4) · 5 minor

  11. Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up, with conditions. Solid empirical comparison of router capacity and prompting for capability routing, but the headline efficiency numbers in the abstract are not independently validated due to circular prompt selection—the paper acknowledges this, but the abstract overstates.

    4 major objections, the first at §3.4, Table 2, b166 · 7 minor

  12. The paper:

    RICapBench: A Fine-Grained Visual Understanding Benchmark Dataset for Vision-Language Models in Remote Sensing

    Computer Science · arXiv:2611.16413

    Xue Wang, Yubin Xu, Kun Tan

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    The passes disagreed. Reader: holds up. Skeptic: agrees with the reader. Referee: major revision.

    Solid benchmark with a clear instance-localization result, but the spatial relation metric rests on an unvalidated parser.

    2 major objections, the first at Section 3.2, Spatial Relationship Parser · 4 minor

  13. The paper:

    Exact Quantum Algorithms for Composed Boolean Functions

    Computer Science · arXiv:2611.16412

    Gagan Chiruvella, Chandrima Kayal, Rajat Mittal and 2 more

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up. Strong paper: general composition theorem for exact quantum query complexity, improved separations, rigorous proofs. Soft spots are openly discussed and minor.

    3 major objections, the first at §4.5, Eq. (4.14) · 6 minor

  14. The paper:

    Flexible Intelligent Metasurfaces for Enhancing Secure MISO Communications

    Systems · arXiv:2611.16411

    Zihao Teng, Zihao Zhao, Jiancheng An and 3 more

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    The passes disagreed. Reader: holds up. Skeptic: holds up only with conditions. Referee: major revision.

    Applies FIM to physical layer security with solid optimization, but the rank-one citation for multi-antenna Eve needs a closer look.

    1 major objection at Section III-A, Lemma 1 · 6 minor

  15. The paper:

    A Lagrangian Floer Theory over the Integers and Finite Fields

    Mathematics · arXiv:2611.16410

    Mohamad Rabah

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up, with conditions. Ambitious extension of integral Lagrangian Floer theory that reads as a plausible blueprint, but the entire argument rests on the author's own unpublished preprints – the paper is not self-contained and cannot be evaluated independently.

    5 major objections, the first at 1.1, Theorem 4.14, Theorem 2.36 · 7 minor

  16. The paper:

    Experience-Handoff Reinforcement Learning for Recursive Self-Improvement

    Computer Science · arXiv:2611.16409

    Doyoung Kim, Byeongguk Jeon, Kailin Zhang and 1 more

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up, with conditions. Single training run per method limits statistical robustness despite careful design.

    2 major objections, the first at Section 4, Table 2, Figure 4, and Appendix H · 7 minor

  17. The paper:

    Gravity and Flight: Learning to Navigate Fine-Grained Remote Sensing Scenes

    Computer Science · arXiv:2611.16408

    Yuheng Zong, Minghua Wang, Xin Zhao and 3 more

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · of version 1 of the paper; version 2 is newer · what was read was not recorded

    The passes disagreed. Reader: holds up. Skeptic: agrees with the reader. Referee: major revision.

    Solid empirical paper with a clean method and honest limitations, but the single-run evaluation leaves the headline improvement unaccompanied by any variance estimate.

    2 major objections, the first at Section G, sentence b615 · 6 minor

  18. The paper:

    Minimal vacuum histories beyond ΛCDM: model comparison within VCDM

    Physics · arXiv:2611.16407

    Antonio De Felice, Eleonora Di Valentino, Luis A. Escamilla and 1 more

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · of version 1 of the paper; version 2 is newer · what was read was not recorded

    Holds up. Solid, transparent model comparison that honestly reports it can't distinguish the alternatives — the nonlinear prescription is a known caveat, not a hidden flaw.

    2 major objections, the first at §II.A, §VI · 5 minor

  19. The paper:

    One-Sided Randomization in Two-Sided Platform Experiments: Which Side Should be Randomized?

    Statistics · arXiv:2611.16406

    Ramesh Johari, Anushka Murthy

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up. Solid paper that convincingly shows that under high market imbalance, short-side randomization dominates both bias and variance, but the central asymptotic regime is restrictive and limits practical generality.

    1 major objection at Section 2.5 and Section 5 (Conclusion) · 4 minor

  20. The paper:

    ReSID: Rethinking Long-Tail Retrieval in Generative Recommendation through Item Reachability

    Computer Science · arXiv:2611.16404

    Muyu Zou, Shunye Tang, Yi Xu and 4 more

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up. Solid, well-executed paper that tackles a real industrial bottleneck in generative recommendation with a practical three-stage fix; the evidence is thorough and the claims hold up.

    2 major objections, the first at Section 4.2, Eq. (2) · 7 minor

  21. The paper:

    Stellar Loci XI: Identifications and Parameter Determinations of Binary Stars by multi-band photometry

    Physics · arXiv:2611.16403

    Yuanchang Wang, Haibo Yuan, Zexi Niu and 4 more

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up. Solid and well-validated method paper — the 65-band SED fitting with MCMC is new for binary detection, and the validation across LAMOST, Gaia NSS, M67, and external catalogs is genuinely thorough. …

    5 minor

  22. The paper:

    Teach with Blinks: Breaking Linguistic Momentum in On-Policy Distillation for Fine-Grained Multimodal Perception

    Computer Science · arXiv:2611.16402

    Tangzheng Lian, Hang Zhang, Guanyu Hu and 4 more

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up. BlinkOPD convincingly fixes linguistic momentum in on-policy self-distillation with a clever eyes-closed reference and first-glance gain, achieving strong data efficiency—recommend acceptance but note limited model family coverage.

    3 major objections, the first at Section 3.2, Appendix G · 5 minor

  23. The paper:

    Probing the extreme quantum violation of the Leggett-Garg inequality

    Physics · arXiv:2611.16401

    Akarshit Baranwal, Vishnu Purushothaman, A.K. Pan

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up. A clear, well-argued corrective to two recent claims of extreme TTB violation; the central argument holds, with a minor philosophical loose end that is honestly flagged.

    2 major objections, the first at §III.C, Eq. (24) and surrounding text · 6 minor

  24. The paper:

    FuGIBench: Evaluating Reliable Preparation in Streaming Video

    Computer Science · arXiv:2611.16400

    Changheng Cai, Zhichao Huang, Song Ren and 8 more

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    The passes disagreed. Reader: holds up. Skeptic: holds up only with conditions. Referee: minor revision.

    FuGIBench defines a clean new evaluation axis for streaming video assistants, but its synthetic videos leave the connection to real-world preparation behaviors unvalidated.

    1 major objection at §4.1, §7 · 4 minor

  25. The paper:

    Saidzoda-Kazakh-EmbeddingGemma-300M: How Far One Epoch of Generated Questions Takes a Small Kazakh Embedder

    Computer Science · arXiv:2611.16399

    Tohir Saidzoda, Madina Abdusalomzoda, Abubakr Abdullozoda and 3 more

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up, with conditions. An unusually honest small-artifact report: a released Kazakh EmbeddingGemma fine-tune plus the only same-code Belebele comparison of the informally shared Kazakh embedders, with a headline gain whose contamination guard is far weaker than its "0 found" suggests.

    2 major objections, the first at Section 4, Section 6, Limitations · 5 minor

  26. The paper:

    Light-Meson Spectroscopy at COMPASS

    Physics · arXiv:2611.16397

    David Sp\"ulbeck (for the COMPASS Collaboration)

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up, with conditions. A competent conference proceedings summary of COMPASS results; the branching ratio is the only genuinely new number, and it carries model dependence the authors themselves flag.

    3 major objections, the first at Sec. 3.1, Eq. (4) and footnote 2 · 5 minor

  27. Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up, with conditions. Solid numerical analysis of a bead-space Cayley integrator for RPMD; the practical dRPMD speedup is plausible but unverified on a real potential.

    2 major objections, the first at Section 5.5, Fig. 4 · 6 minor

  28. The paper:

    TriGen: A Generative Paradigm for Trigger-Induced Recommendation

    Computer Science · arXiv:2611.16395

    Muyu Zou, Haibo Xing, Shunye Tang and 4 more

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up, with conditions. TriGen is a solid industrial paper that defines a new setting (trigger-induced generative recommendation) and backs it with a live deployment. …

    4 major objections, the first at Appendix D, Table 1 · 8 minor

  29. The paper:

    Exceptional point excitation by a plunging particle

    Physics · arXiv:2611.16390

    Xia-Yuan Liu, Liang-Bi Wu

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up, with conditions. A clean, internally consistent demonstration that a plunging particle excites the EP ringdown and that conventional templates can be Bayes-favored yet wrong by 21% in the damping rate. …

    3 major objections, the first at §IV.B, Fig. 5, and surrounding text · 11 minor

  30. The paper:

    On Energy-Based Modelling for Fluidic Excitable Neurons

    Systems · arXiv:2611.16389

    Robin Wroblowski, Brent Michiels, Benjamin Gorissen and 1 more

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up, with conditions. Solid bond-graph translation of excitability into pneumatics, but the load-bearing lumped approximation is untested and the core EI model is from an unpublished preprint.

    2 major objections, the first at Section III, 'Lumped pneumatic regime'; Section V.A · 5 minor

  31. The paper:

    Strategically Robust Multi-Agent Control

    Mathematics · arXiv:2611.16388

    Boris Velasevic, Nicolas Lanzetti, Eric Mazumdar

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up, with conditions. A well-executed paper that introduces a clean equilibrium concept interpolating between Nash and H-infinity control, with solid proofs and interesting numerics; the main weakness is the lack of reproducible code for the experiments.

    no objections raised

  32. The paper:

    Quantum geometry drives orbital paramagnetism in superconducting fluctuations

    Physics · arXiv:2611.16387

    Zi-Ting Sun, Ying-Ming Xie, Chunli Huang and 1 more

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    The passes disagreed. Reader: holds up. Skeptic: holds up only with conditions. Referee: accept.

    A careful theory paper whose central decomposition holds up within its stated assumptions; the kagome paramagnetic prediction is conditional on non-minimal transverse couplings the authors flag but do not fix.

    5 minor

  33. The paper:

    Remember What Changes the Odds: Critic-Gated Event Memory for Long-Horizon Manipulation

    Computer Science · arXiv:2611.16386

    Xuanzhuo Liu, Chuankang Li, Maoqing Yao and 1 more

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    An outcome-grounded event-memory principle with a correct martingale bound and a strong synthetic diagnostic; the real-data section tests a different, stronger signal than the CGEM gate writes on, so the hardware bridge is still open.

    the referee pass did not run

  34. Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up, with conditions. Useful B-field map and clean Ricean depolarization measurement for Serpens Main, but the RAT alignment-loss interpretation needs a partial-correlation test before it carries the weight the abstract puts on it.

    3 major objections, the first at §4.4 and Figure 13 · 4 minor

  35. The paper:

    Beyond Compressing Everything: Contrastive Memory Learning for Selective Context Compression

    Computer Science · arXiv:2611.16384

    Sikui Zhang, Haowei Liu, Ziyun Gan and 9 more

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up, with conditions. Worth a careful referee pass: the same-context hard negatives and bidirectional mask form a real, well-ablated contribution, but the full-context comparison is overclaimed and the mean-pooled positive target deserves scrutiny.

    3 major objections, the first at §3.2.1, Table 1 · 4 minor

  36. The paper:

    Continuous-Time Random Walks with Internal Stochastic Renewal Process

    Physics · arXiv:2611.16383

    Boris Z. Huang

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up, with conditions. A CTRW with infinitely many internal states that increment or reset; the kernel algebra is sound and the three applications are new, but the spectral eigenrelation (2.7) is not verified …

    3 major objections, the first at §2, Eq. (2.7) and §3, Eq. (3.2) · 5 minor

  37. The paper:

    RADIANT: Redundancy-Aware Information Allocation for Universal Multi-Modal Fusion

    Computer Science · arXiv:2611.16382

    Shenghai Wu, Yue Lin

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up, with conditions. A well-executed fusion framework with a clear design philosophy and strong empirical results, but the universality claim is stretched and the headline detection result lacks statistical rigor.

    4 major objections, the first at Abstract; Section 4.2; Table 7 · 7 minor

  38. The paper:

    Dynamic Tunneling a Prior Phenomenon to Intermediate Resistive States in Defect Engineered Mn-SrTiO3 RRAM Devices

    Physics · arXiv:2611.16380

    Gowtham Velpandi, Ishaan Tushir, Durairaj Murugan and 2 more

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up, with conditions. Solid, referee-worthy RRAM device work where the headline tunneling claim is a fitted hypothesis, not an observation — the abstract outruns the authors' own hedging.

    5 major objections, the first at §4, Eqs. (7)–(8); Table S2 · 6 minor

  39. The paper:

    p-adic Height Pairing through Higher Class Field Theory: Part I

    Mathematics · arXiv:2611.16379

    Yuhui Yao

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    The passes disagreed. Reader: holds up, with conditions. Skeptic: gave no verdict. Referee: reject.

    Genuinely new idea, honestly written, but the p-local reciprocity - exactly where the construction is new - is asserted, not derived, so Theorem 1.1 is not proved as the manuscript stands.

    5 major objections, the first at §2, sentence b44 · 3 minor

  40. The paper:

    Evidence-Governed Adoption of Self-Generated Updates in Autonomous AI Agents

    Computer Science · arXiv:2611.16378

    Subodh Lonkar

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up. Correct, honest, and modest: a clean statistical gate for agent-update adoption whose load-bearing assumption the authors themselves expose.

    This review says the paper's mathematics is wrong. Pith did not check that against the paper's source.

  41. The paper:

    Universal optimality on finite Lee tori

    Mathematics · arXiv:2611.16377

    Tao Zhang

    Pith says

    Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded

    Holds up, with conditions. Settles the Lee-torus checkerboard conjecture for all even side lengths with a genuinely new LP strengthening; the odd-torus classification's long coefficient chains survive independent checking.

    1 major objection at Lemma 5.2, displayed reduction after Eq. (5.1) · 4 minor