Pith.

Every reviewed paper

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

498,816

papers reviewed

live · read 16 Nov 2026, 02:14 UTC

Narrow the list

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

  1. The paper:

    Back to Pixels: Lifting Latent Video Diffusion into Pixel Space

    Computer Science · arXiv:2611.17336

    Yanke Song, Shuo Chen, Chris Rockwell and 9 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.

    A strong, well-evaluated recipe paper: lifting a pretrained latent video model into frame-wise pixel space works end-to-end and buys real resolution, editing, and latency benefits; the linear-lift mechanism is unvalidated but not load-bearing for the measured claims.

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

  2. The paper:

    Long-lived light-driven bimagnons at finite momentum in NiO

    Physics · arXiv:2611.17335

    Xian P. Yang, Christopher T. Parzyck, Martin Cross and 21 more

    Pith says

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

    No verdict was given on this paper.

    4 major objections, the first at Fig. 2c and associated text · 6 minor

  3. The paper:

    Unbundling Multimodal Vision Encoding

    Computer Science · arXiv:2611.17334

    Ellis Brown, Chris Hoang, Shengyi Qian and 9 more

    Pith says

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

    Holds up, with conditions. Careful controlled study unbundling vision encoder design choices; the main claims hold up within scope, but the fixed data mixture is a real load-bearing limitation that the authors acknowledge but do not resolve.

    4 major objections, the first at §2.2, Appendix A.4, §8 · 8 minor

  4. The paper:

    Discrete Time Crystals from Anti-Unitary Floquet Driving and Dynamical Freezing

    Physics · arXiv:2611.17333

    Anirban Das, Asmi Haldar, Arnab Das

    Pith says

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

    A solid, well-executed paper that delivers what it promises: a practical, parameter-free method for estimating the local density of states in disordered systems. …

    the referee pass did not run

  5. The paper:

    TTB: Test-time Baking for Efficient Rendering of Decoder-only View Synthesis Models

    Computer Science · arXiv:2611.17331

    Chin-Yang Lin, Ryo Hachiuma, Jaesung Choe and 6 more

    Pith says

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

    Holds up, with conditions. TTB is a solid engineering contribution that trades reconstruction time for rendering speed, but the 'on par' quality claim is overstated for the primary variant.

    4 major objections, the first at Table 1, V=64 row · 8 minor

  6. The paper:

    ArchSpace: Recording, Reviewing, and Referencing LLM Architecture Research

    Computer Science · arXiv:2611.17332

    Jie Wang, Tong Zhu, Jiahao Kuang and 8 more

    Pith says

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

    The scaling-law work is solid and unusually honest about its limits; the platform is a credible proposal, not yet a system, and the paper deserves a serious referee.

    the referee pass did not run

  7. The paper:

    ProtoVis-Nav: Prototypical Visual Imagination for Visual-Spatial Aligned UAV Navigation

    Computer Science · arXiv:2611.17330

    Guangyuan Sui, Yunzhong Hou, Chi Harold Liu

    Pith says

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

    Holds up, with conditions. A solid, well-evaluated paper that introduces a cheap local-imagination trick for UAV VLN; worth refereeing, with minor requests for error bars and view-selection ablations.

    5 major objections, the first at Table 1 · 9 minor

  8. The paper:

    Internalizing Investment Mandates: A Neurosymbolic RL Approach to Portfolio Allocation

    Computer Science · arXiv:2611.17329

    Zeynep Y{\i}ld{\i}z, Satchit Chatterji, Ercan Atam and 1 more

    Pith says

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

    Holds up, with conditions. PLPG-MC is a clean adaptation of neurosymbolic constraints to continuous actions, and the experiments are honest about limitations, but the λ sweep may have been done on the test set, which is a load-bearing methodological gap.

    2 major objections, the first at §4.3, Figure 3 · 7 minor

  9. The paper:

    A Dark-Sector Mass Flip-Flop Mechanism for the ANITA-IV Events

    Physics · arXiv:2611.17328

    Dibya S. Chattopadhyay, Carlos A. Arg\"uelles, Vedran Brdar

    Pith says

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

    Clever BSM fix for ANITA–IceCube, but the angular fit is marginal and the PUEO prediction is not independent.

    the referee pass did not run

  10. The paper:

    Decoupling Dynamics: A Tale of Two Limits

    Physics · arXiv:2611.17327

    Andrea Antinucci, Oscar Lewis

    Pith says

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

    Holds up, with conditions. The paper gives a clean quantum explanation for how the SymTFT changes under dynamical gauging of a continuous U(1) symmetry, resolving a tension in the literature. …

    5 major objections, the first at §3.1 · 10 minor

  11. The paper:

    A hyperbolic area proof of the Kahn-Lyubich Covering Lemma and its generalization

    Mathematics · arXiv:2611.17326

    Kostiantyn Drach

    Pith says

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

    A clean, self-contained hyperbolic-area proof of the Kahn–Lyubich Covering Lemma, plus an extension to infinite-degree maps. The proof is transparent and the reliance on an unpublished lemma is acknowledged. The paper is a solid technical contribution that deserves peer review.

    the referee pass did not run

  12. The paper:

    Bound States of Two D0-Branes and the Three-Graviton Coupling

    Physics · arXiv:2611.17325

    Sergio Sanjurjo, Themistocles Zikopoulos

    Pith says

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

    Holds up, with conditions. The paper delivers a genuine numerical solution for the SU(2) BFSS ground state and a clean analytic derivation of the three-graviton coupling; the short-range assumption in the analytic argument is …

    1 major objection at Section III, Eqs. (21)–(23); Appendix E (b203, b495) · 6 minor

  13. The paper:

    K2V: Generative Keyframing For Physically Plausible Video Generation

    Computer Science · arXiv:2611.17324

    Felix O'Mahony, Daniel Olmeda Reino, Roberto Cipolla and 1 more

    Pith says

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

    Holds up, with conditions. A no-training keyframe scaffold that moves Physics-IQ by 22.8 points, with a clean ground-truth control isolating the mechanism — genuinely worth refereeing, though the mechanism is single-backbone and the PhyGenBench SOTA is anchored to a self-run competitor.

    4 major objections, the first at §3.2.2, Eq. (3) · 5 minor

  14. The paper:

    AgentSSL: Can MLE Agents Leverage Unlabeled Data?

    Computer Science · arXiv:2611.17323

    Ethan Lin, Akanksha Sarkar, Ziang Liu and 2 more

    Pith says

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

    Holds up, with conditions. The selection-bottleneck finding and 17k-program benchmark are the real contribution; the '8 of 12' headline is an oracle claim calibrated against a baseline pool that isn't the strongest human engineering.

    5 major objections, the first at Table 1, Section 5 · 9 minor

  15. The paper:

    Ensemble phase transitions in measurement-unitary hybrid quantum circuits

    Physics · arXiv:2611.17322

    Saptarshi Mandal, Sthitadhi Roy

    Pith says

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

    Convincing demonstration that trajectory and projected ensembles undergo phase transitions at the MIPT, with clean analytics for Clifford and numerics for Haar circuits. The paper is solid within its stated scope.

    the referee pass did not run

  16. The paper:

    BrickFlow: Connectivity-Guided Brick Reconstruction

    Computer Science · arXiv:2611.17321

    Peter Kulits, Cordelia Schmid

    Pith says

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

    Holds up, with conditions. First inventory-conditioned LEGO reconstruction with a clever test-time guidance, but the guidance benefit is conflated with snapping and absolute validity is low.

    2 major objections, the first at §3.7, §B, Tabs. 1 and 3 · 5 minor

  17. The paper:

    Learning When Diffusion Language Models Should Regenerate

    Computer Science · arXiv:2611.17320

    Minhae Oh, Jeonghye Kim, Jinwoo Koo and 4 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.

    4 major objections, the first at §4.2, Table 1 · 7 minor

  18. The paper:

    The Cauchy problem for skew-torsion-parallel spinors with applications to supergravity

    Mathematics · arXiv:2611.17319

    C. S. Shahbazi

    Pith says

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

    Holds up, with conditions. Solves the Cauchy problem for skew-torsion-parallel spinors via isotropic parallelisms; the compact rigidity theorem has a gap in the leafwise Hodge projection for dense leaves.

    4 major objections, the first at Lemma 4.38 · 8 minor

  19. The paper:

    Spatial Foundry: A Multimodal Dataset for Spatial Intelligence

    Computer Science · arXiv:2611.17318

    Ethan Lin, Yang Song

    Pith says

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

    Holds up, with conditions. A well-scoped dataset paper that describes a useful resource but ships no data, code, or validation. The concept is sound; the release is what matters.

    4 major objections, the first at §4.4–4.5, Eq. (6) and (9) · 8 minor

  20. The paper:

    C1,α Regularity of W2,2 Q-Valued Maps with Small Laplacian

    Mathematics · arXiv:2611.17317

    Mattia Luchese

    Pith says

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

    Holds up, with conditions. Serious, technically rich paper that proves a new C^{1,α} regularity criterion for W^{2,2} Q-valued maps with small Laplacian, with applications to stationary maps and stable minimal immersions. …

    5 major objections, the first at §2.1, Lemma 1 · 9 minor

  21. The paper:

    CoReMem: Cost-Internalizing Memory for Successive Coding Tasks with Prospective Valuation

    Computer Science · arXiv:2611.17316

    Peiwen Li (Yale University), Harshit Verma (Yale University), Sophia Zhou (Yale University) and 6 more

    Pith says

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

    Holds up, with conditions. CoReMem is a serious paper with a clean method for valuing memory on successive coding tasks, but the headline comparison on ChainSWE bundles post-training with memory and the paper does not isolate the memory-specific gain on that benchmark.

    4 major objections, the first at Table 1 and §4.2 · 7 minor

  22. The paper:

    Countable dynamical systems with a dense orbit

    Mathematics · arXiv:2611.17315

    Jhon P\'erez-Remolina, Carlos Uzc\'ategui-Aylwin

    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.

    Clean classification result for countable dynamical systems with a dense orbit; the gluing construction is the real contribution, but the proof of Theorem B(ii) for |X^{(2)}| > 1 has a gap that the stress-test note correctly identifies.

    3 major objections, the first at Section 6, Lemma 6.2 and its application in Theorem 7.1(ii) · 9 minor

  23. Pith says

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

    Holds up, with conditions. A genuinely useful, carefully controlled demonstration that MAML's single-step adaptation updates factorize additively by symbolic component — the stress test's additivity worry is mostly answered by the controls, but the …

    4 major objections, the first at §3.2, §4.1, and the skeptic's concern · 8 minor

  24. The paper:

    Sketching Language Models for Context Attribution

    Computer Science · arXiv:2611.17313

    Christopher Liaw, Aranyak Mehta, Grigoris Velegkas and 1 more

    Pith says

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

    Holds up, with conditions.

    4 major objections, the first at §6, Figure 1(b), and Appendix B · 7 minor

  25. The paper:

    The Tournament Equilibrium Set Is PSPACE-Complete, Even under Hereditary Uniqueness

    Computer Science · arXiv:2611.17312

    Ethan Dickey, Athina Terzoglou

    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.

    Closes the TEQ membership gap with a genuinely new reduction; the anchor-deletion stress-test concern does not survive contact with the paper.

    4 major objections, the first at §4.4, Lemma 4.10, and §A.3, Lemma A.3 · 18 minor

  26. The paper:

    Integer Programming with Polyhedral Constraints on the Image

    Computer Science · arXiv:2611.17311

    Anubhav Dhar, Tim Randolph, Karol W\k{e}grzycki

    Pith says

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

    Holds up, with conditions. A solid, well-scoped paper that cleanly solves a natural open problem in combinatorial geometry, with a clear proof and no obvious holes. The main result is new and the exposition is unusually honest about limitations.

    4 major objections, the first at §3.2, Corollary 4 · 10 minor

  27. The paper:

    Quantum Metrology and Quantum Battery in a Hyperfine-Coupled Spin System

    Physics · arXiv:2611.17310

    Omar Bachain, Mohamed Amazioug, Rachid Ahl Laamara

    Pith says

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

    A solid, fully analytic two-spin worked example; the battery half is clean, but the metrology section asserts the load-bearing weak-commutativity condition without showing the calculation, and the promised dephased-state comparison is missing.

    the referee pass did not run

  28. The paper:

    Classical Energies Learn to Reason 1,000x Faster than Looped Models

    Computer Science · arXiv:2611.17309

    Benjamin Hoover, Hugh Blayney, James C. R. Whittington

    Pith says

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

    Holds up, with conditions. CRAM is a clean, well-executed paper that shows a learned mean-field energy solves simple constraint puzzles orders of magnitude faster than looped baselines. …

    5 major objections, the first at App. C.2, Table 2, Fig. 2a · 7 minor

  29. The paper:

    Separations Between Submodular-Like Graph Widths

    Computer Science · arXiv:2611.17307

    Karl Bringmann, Egor Gorbachev

    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 paper, main claims hold, but the MILP-dependent separations are not formally verified and the paper acknowledges this honestly.

    2 major objections, the first at Section 8.2, MILP for m-width · 22 minor

  30. The paper:

    Statistics of Fock-space correlation lengths in the many-body localised phase

    Physics · arXiv:2611.17306

    Bibek Saha, David E. Logan, Sthitadhi Roy

    Pith says

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

    Holds up, with conditions. A clean, internally consistent theory of Fock-space correlation-length distributions in the MBL phase, with strong numerical support; the singularity-spectrum extension is plausible but rests on an unvalidated self-averaging assumption.

    3 major objections, the first at §5.1, Eq. (24) and surrounding text (b99, b180) · 9 minor

  31. The paper:

    Where LLM Agents Get Stuck Debugging Databases: Access More Than Reasoning

    Computer Science · arXiv:2611.17304

    Gustavo Pinto

    Pith says

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

    Holds up, with conditions. First genuine field report of a production-debugging agent, honestly limited, but the 'access over reasoning' headline is a single-author interpretation of four mixed cases, not an observation.

    3 major objections, the first at §5, Table 2; §7 [b213] · 6 minor

  32. The paper:

    Towards a Theory of Learnability via Debate

    Computer Science · arXiv:2611.17303

    Tosca Lechner, Vinayak Pathak, Sivan Sabato and 1 more

    Pith says

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

    Holds up. First real learning-theoretic treatment of debate as a learning problem, with a clean dimension-based characterization and an uncomfortable noise-brittleness result; the math holds up and the soft spots are cosmetic.

    11 minor

  33. The paper:

    DiSKO: Distribution Steering for Nonlinear Systems using the Koopman Operator

    Systems · arXiv:2611.17301

    Dimitrios Moustroufis, Panagiotis Tsiotras

    Pith says

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

    Holds up, with conditions. Solid integration of Koopman and DR-CS with a new initial-ambiguity-bound theorem; the data-coverage premise for the disturbance ambiguity set is the main unresolved issue, and Table II shows exactly the failure mode.

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

  34. The paper:

    JWST Rocky Worlds DDT Observations of LTT 1445 A b and c in Secondary Eclipse Using MIRI/F1500W

    Physics · arXiv:2611.17300

    Jegug Ih, Shubham Kanodia, Nicholas Wogan and 15 more

    Pith says

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

    Holds up. Solid eclipse paper: new MIRI/F1500W data, clean analysis, honest about planet c's grazing geometry, and a useful CO2 upper limit for b. The Shoreline inference is modest but fair.

    3 major objections, the first at §4.1, Table 2, and §7.2.3 · 8 minor

  35. The paper:

    Consistent Video-to-Video Translation via Explicit Correspondences

    Computer Science · arXiv:2611.17298

    Gaurav Parmar, Zhengqi Li, Richard Zhang and 4 more

    Pith says

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

    Holds up, with conditions. A clean, well-ablated method for long-range consistency in video-to-video translation; the core idea is genuinely new and the experiments are thorough, but the lack of error bars and threshold tuning on the evaluation benchmarks are real concerns.

    3 major objections, the first at §4, Tables 1–3 and 5 · 4 minor

  36. The paper:

    Goal Encoding in LLMs

    Computer Science · arXiv:2611.17297

    Arnab Sen Sharma, Eric Todd, Sheridan Feucht and 7 more

    Pith says

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

    Holds up, with conditions. A serious, well-executed paper that localizes goal-like representations in LLM attention heads and proposes a non-recurrent persistence mechanism; the central claims hold up in code settings, but generalization beyond code is thin and the NRGS account is supported only partially.

    5 major objections, the first at §3.1, Appendix C · 8 minor

  37. The paper:

    Dynamic k-Median with Worst-Case Update Time and Recourse

    Computer Science · arXiv:2611.17296

    Chenglin Fan

    Pith says

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

    Holds up, with conditions. This is a serious and likely correct resolution of the amortized-to-worst-case gap for dynamic k-median, and the Step 6 concern raised in the stress test does not hold up against the actual proof.

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

    4 major objections, the first at §8.3, Lemma 8.7 and surrounding text · 17 minor

  38. The paper:

    Metal Sandbox: Isolated Execution and Command Scheduling for GPU Coding Agents on Apple Silicon

    Computer Science · arXiv:2611.17294

    Haozhe Wang, Lina Chen, Fanqi Wan and 5 more

    Pith says

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

    Holds up, with conditions. An honest, well-measured systems paper whose live gains are solid; the replay headline is directionally safe but rests on one unquantified assumption about what the measured tool waiting is made of.

    2 major objections, the first at Sections 5.6-5.7 and Appendix B · 5 minor

  39. The paper:

    Three-ends Furstenberg estimates and approximate rectangles

    Mathematics · arXiv:2611.17293

    Ciprian Demeter, William O'Regan

    Pith says

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

    Holds up, with conditions. New three-ends incidence estimate with genuine applications; sharpness claim for exponent 5/4 is not fully justified.

    3 major objections, the first at Section 1.1, sharpness discussion · 13 minor

  40. The paper:

    Radiation, Thermochemistry, and Stellar Feedback in DISPATCH

    Physics · arXiv:2611.17290

    William Lake, Paolo Padoan, {\AA}ke Nordlund 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, with conditions. A thorough, honest methods paper that validates each module separately but never runs the full coupled system; the central claim is extrapolated.

    3 major objections, the first at Abstract; Sections 4.4–4.8 · 5 minor

  41. The paper:

    Sharp Uniform Fisher Information Bounds from Projective Designs

    Physics · arXiv:2611.17288

    Zhiyao Yang, Bowen Xiao, Huangjun Zhu

    Pith says

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

    Holds up. Sharp Fisher guarantees for projective designs, with a useful O(d)-outcome construction. Solid paper, send to referee.

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

  42. The paper:

    Negative-Only Policy Optimization for One-Sided Verifiable Rewards

    Computer Science · arXiv:2611.17287

    Jiacheng Xu, Shuo He, Fuxiang Zhang and 2 more

    Pith says

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

    A genuinely useful formalization of one-sided verifiability with a sensible negative-only update; the broad empirical claim needs repeated seeds and code before I trust the point margins.

    the referee pass did not run