arXiv subjects

Kevin Guan

Publications and source records attributed to Kevin Guan.

6 recordsLinked to original sources

arXiv2026

The transversal achievement game on a square grid

In the transversal achievement game on the $n\times n$ board, two players alternately claim cells, and the first to own a transversal---a set of $n$ cells of which no two share a row or column---wins. Ran{\dj}elovi\'c showed that the first player wins for every $n\ge4$, while the game is a draw for $n=2,3$. We give an independent proof that the first player wins for $n\ge4$ that additionally establishes a bound on the length of the win: the given strategy forces a win by ply $2n+3$, i.e.\ on the first player's $(n+2)$-nd move, for every $n\ge4$. The proof yields a strategy that is fully determined by a fixed rule on the current position and can thus be implemented directly. We isolate the use of the hypothesis $n\ge4$ to two steps in the analysis, explaining why the argument fails at $n=3$. An exhaustive computational search implementing the strategy verifies it against every legal defense for $n=4,5,6$, confirming both the strategy's validity and that the $2n+3$ bound is attained in these cases. The main theorem has also been formalized and machine-checked in Lean 4.

math.CO↗

arXiv2026

MDArena: Evaluating Coding Agents on Realistic Molecular Dynamics Workflows

Accelerating scientific discovery is among the most consequential applications of AI, and computational biomolecular simulation stands out as a particularly promising target within this broader effort. Coding agents promise to automate significant portions of this workflow, yet their reliability on realistic molecular dynamics (MD) tasks remains poorly characterized. To address this issue, we introduce MDArena, a benchmark of 50 containerized tasks drawn from active biomolecular simulation projects, spanning 29 molecular systems and 14 broad research protocols, including trajectory analysis, complex system preparation, free-energy protocols, and enhanced sampling. We evaluate six model/harness configurations spanning Codex and OpenCode. Among the evaluated configurations, Codex GPT-5.5 at extra-high reasoning effort performs best, reaching 24/50 Strict-Pass@1 successes (48%), followed by Codex GPT-5.5 Medium with 21/50, and OpenCode Gemini Flash 3.5 with 20/50. Average correctness and process rewards are substantially higher than strict success rates across all configurations, indicating that agents frequently make meaningful partial progress but fail on the fine-grained details required for reproducible scientific workflows. Hard tasks remain largely unsolved, particularly membrane-protein system preparation and alchemical free-energy setup, both unsolved or near-unsolved by every evaluated configuration. MDArena thus exposes a substantial gap between the usefulness of coding agents as supervised assistants and their reliability as autonomous MD researchers, while providing a reproducible and extensible platform for tracking progress toward closing it.

physics.chem-ph↗

arXiv2026

Latent States in Neural Networks: Recovering the Temporal Structure of Drifting Data from Model Weights

A temporally drifting data stream may pass through discrete regimes rather than changing continuously. We ask whether such regimes are recoverable from the weights of models trained on the stream, using a hidden Markov model (HMM) fit to the chronologically ordered trajectory of those weights. We study this question in two domains known to drift over time: multimodal misinformation detection, using the Fakeddit dataset; and sentiment analysis, using the Yelp dataset. We train classifiers on consecutive temporal windows and fit an HMM to the trajectory of their aligned weights, recovering latent states that partition each timeline into coherent phases. On both datasets, classifiers generalize better to data from windows sharing the state of their training window than to windows across state boundaries. This within-state transfer advantage survives a control for temporal proximity and modestly exceeds the advantage recovered by a naive partition into contiguous states of equal size. Although the states are estimated solely from model weights, they correlate more strongly with shifts in the data's class distribution than with the weight-space geometry used to estimate them. After class divergence and lag are residualized out, the within-state advantage exceeds its permutation null on both tasks, indicating that the states recover structure relevant to transfer beyond the data distribution. Every effect replicates on both tasks but is attenuated on Yelp, whose label distribution is more temporally stable.

cs.LG↗

arXiv2026

AfriSUD: A Dependency Treebank Collection for Evaluating Models on African Languages

Despite their linguistic diversity and global significance, African languages remain underrepresented in research and resources to support NLP. We aim to bridge this gap by introducing AfriSUD, the first large-scale collection of syntactically annotated treebanks for nine diverse African languages spanning major language families and regions across Sub-Saharan Africa. Using the Surface-Syntactic Universal Dependencies (SUD) framework, our community-led effort provides high-quality, native-speaker verified data that capture typological key features such as agglutination and tone. We evaluate a range of models on AfriSUD for part-of-speech tagging and dependency parsing including non-transformer baselines, multilingual pretrained encoders, and LLMs. Our results reveal a significant syntax gap, where models still show clear limitations across the nine languages, suggesting that existing architectures may not fully capture the structural diversity of African-language syntax.

cs.CL↗

arXiv2026

Dependency Parsing Across the Resource Spectrum: Evaluating Architectures on High and Low-Resource Languages

Transformer-based models achieve state-of-the-art dependency parsing for high-resource languages, yet their advantage over simpler architectures in low-resource settings remains poorly understood. We evaluate four parsers---the Biaffine LSTM, Stack-Pointer Network, AfroXLMR-large, and RemBERT---across twelve typologically diverse languages, with a focus on low-resource African languages. We find that the Biaffine LSTM consistently outperforms transformer models in low-resource regimes, with transformers recovering their advantage as training data increases. The crossover falls within a resource range typical of treebanks for under-resourced languages. Morphological complexity (measured via MATTR) emerges as a significant secondary predictor of transformers' relative disadvantage after controlling for corpus size. These results indicate that the Biaffine LSTM may be better suited for syntactic tool development in low-resource regimes until sufficient annotated data is available to leverage the representational capacity of pre-trained transformers.

cs.CL↗

arXiv2025

A Generic and Efficient Python Runtime Verification System and its Large-scale Evaluation

Runtime verification (RV) now scales for testing thousands of open-source Java projects, helping find hundreds of bugs. The popular Python ecosystem could use such benefits. But, today's Python RV systems are limited to a domain or specification logic, or slow. We propose PyMOP, a generic, extensible, and efficient RV system for Python. PyMOP supports five logics, implements five existing monitoring algorithms, ships with 73 API specs of Python and widely-used libraries, supports three instrumentation strategies, and users can easily add more of these. On 290,133 unit tests in 1,463 GitHub projects, we find mainly that (i) the default monitoring algorithm for Java is often not the fastest for Python; (ii) PyMOP is up to 1,168.3x faster than two recent dynamic analysis systems; and (iii) 44 of 121 bugs that PyMOP helped find so far were fixed by developers. PyMOP's generality and efficiency position it well as an excellent platform for the next advances on RV for Python.

cs.SE↗