Evolve Quick Reviews
Public Paper Reviews
Structured preliminary reviews of recent open-access papers, generated and continuously updated by the Evolve analysis engine. Each review condenses what the paper claims, how solid the evidence is, and the questions a rigorous peer reviewer would ask — English analysis with abstracts in your language. These are independent AI analyses, not peer review; the canonical record always lives at the publisher.
Beyond Confidence: Stability-Aware Test-Time Adaptation for LLM Reasoning
Bincheng Gu, Min Gao, Zongwei Wang, Yibing Bai et al.
TASCO adds stability regularization to test-time confidence optimization with solid gains, but transductive test-set fitting and uneven gains temper claims.
Agent-Integrated Software: Interaction Contracts and Continuous Assurance
Shengcheng Yu, Chunrong Fang, Zhenyu Chen
Perspective formalizing agent-in-app delegation as interaction contracts plus versioned assurance; careful and honest, but unimplemented and unvalidated.
Improving the Sensitivity of Gravitational Wave Detection with Weighted Conformal Prediction
Ann-Kristin Malz, Gregory Ashton, Nicolo Colombo
Weighted conformal prediction restores CP coverage for GW pipeline combination under shift on mock data; sensitivity gains trade against 6x false positives.
From Queries to Narratives: Cultural Heritage Data Stories for Knowledge Graph Exploration and Quality Assessment
Tabea Tietz, Torsten Schrade, Etienne Posthumus, Linnaea Söhn et al.
Presents LODEON data stories for exploring NFDI4Culture-KG via live SPARQL narratives; promising system but evidence is qualitative, AI accuracy unquantified.
Cross-Lingual Clinical Annotation Projection as Constrained Text Generation: A Six-Language Study
Álvaro Rey-Blanes, Francisco J. Moreno-Barea, Francisco J. Veredas
Constrained LLM projection beats prior SOTA across 18 language-entity settings (Strict F1 0.9201); solid, but no significance tests or contamination analysis.
TransClean: A Benchmark for Detecting and Extracting Clean Translations from Large Language Model Outputs
Shenbin Qian, Yves Scherrer
First benchmark for extracting clean translations from noisy LLM outputs; thorough analysis, but LLM-annotated silver labels raise circularity concerns.
Deep-Fake CAPTCHA: Mitigating Next-Generation Social Engineering Attacks
Guy Frankovits, Lior Yasur, Fred M. Grabovski, Yisroel Mirsky
Extends DF-CAPTCHA from voice to video: active challenges beat passive detection, but both modalities use a single fake model and no adaptive attacker.
SWRouter: Similarity-Contractive Window Routing for Multi-Turn Large Language Model Conversations
Yu Wang, Yuchen Li, Rui Kong, Xinran Chen et al.
Proposes similarity-window routing for multi-turn LLMs; reports 16.26/8.22-point ShareGPT+MTBench gains. Promising, but P_router inconsistency, test-set-tuned threshold, no significance testing demand verification.
Calibration-Aware Uncertainty Cascades for Efficient Heterogeneous Model Collaboration
Yilin Zhang, Han Jiang, Cai Xu, Ying Liu et al.
CAUC: training-free calibrated cascades; simple, theoretically sound, consistent gains — but single-run evidence and an unverified 47% claim warrant caution.
RouteRepair: Instance-Level Failure Diagnosis and Targeted Repair in LLM-Based Automated Heuristic Design for Routing Optimization
Binghao Ji, Di Huang, Jiahui Fang, Zhiyuan Liu
RouteRepair adds instance-level failure diagnosis and targeted repair to LLM-evolved routing heuristics, with promising GLS/CVRP/ACO gains and honest, but single-seed, evidence.
From Document Silos to Process Intelligence: A Multi-Layer Knowledge Graph for CMC Process Development
Reza Amirmoshiri, Faryad Sahneh, Yasser Jangjou
Sanofi's dual-layer CMC knowledge graph shows rigorous lexical-layer eval (95% T1/85% T2, 505 Qs), but the domain layer and router remain under-evaluated.
LLMs as Post-hoc Auditors of Physiological Plausibility in Symbolic Regression: A Clinician-Evaluated Case Study
Jorge López-Varela, J. Ignacio Hidalgo, José-Manuel Muñoz, Omar Costilla-Reyes et al.
Case study: LLMs help rank symbolic regression models under clinician oversight, but flawed interpretations and a 3-rater co-author panel limit the evidence.
Hologram Representation via Quadratic Phase Gaussian Splatting
Haolong Wang, Yicheng Zhan, Kaan Akşit, Simeng Qiu
Adds learnable quadratic-phase curvature to complex Gaussians for holograms; small but consistent gains (+0.19/+0.33 dB), simulation-only, single baseline.
Investigating catastrophic forgetting in sound event classification
Riccardo Casciotti, Annamaria Mesaros
Claims frozen features plus a dynamic head best curb forgetting in incremental sound classification; convincing, but baselines and variance reporting are thin.
Prevalence Determines Precision:Silent Contamination in Detector-Defined Datasets
Jia Huang, Yankai Wan, Yangjun Ou
Claims dataset precision is set by pool prevalence via Bayes; futures case study shows +422% naive transfer error — sound and honest, but scope is narrow.
Extending SMT Solving with Non-Ground Clause Learning
Yasmine Briefs, Christoph Weidenbach
ICLF: a calculus unifying instantiation-based SMT, CDCL(T) and non-ground clause learning; soundness, non-redundancy and simulation proofs, no implementation.
Structural priors for data-efficient language learning
Yana Veitsman, Jonas Mayer Martins, Jonathan Lautenschlager, Lisa Beinborn
Symbolic pre-pretraining lowers multilingual LM loss (~60% token budget) but not linguistic benchmarks; well-controlled; correlation rests on ~8 medians.
Flexible and Interpretable Accent Distance Measurements
Charles McGhee, Mark J. F. Gales, Kate M. Knill
Articulatory features + optimal transport give interpretable, unpaired accent distances; promising on VCTK, but no significance tests or broad validation.
ReGround: Grounding Reviewer Comments in Multimodal Evidence
Serwar Basch, Lizhen Qu, Iryna Gurevych
Rebuttal-derived dataset grounds reviewer comments in multimodal paper evidence; solid QA and analysis, but author-cited-only gold caps score meaning.
The Convention Gap: Towards Measuring Implicit Communication in Cooperative AI Evaluation
Makoto Fukushima, Hua-Dong Xiong, Ehsan Moradi Pari
Proposes an exactly computable 'convention gap' for implicit communication in Hanabi; strongly validated via OBL dose-response; cross-dataset confound remains.
Continuous Learning of Gravity Field Irregularities Around Small Bodies via Neural Hamiltonian ODEs
Giacomo Acciarini, Dario Izzo
Claims tracking-only neural Hamiltonian ODE gravity models beat harmonics inside the Brillouin sphere; synthetic but well-controlled evidence is persuasive.
Published Unlearning Numbers Move Per Checkpoint, and Not Because the Removed Data Survives: An Audit of 263 Released Batch-Normalized Checkpoints
Junlong Shen Xingyu Li
Audits 263 released unlearning checkpoints; refitting BN state moves numbers per checkpoint but not via removed data; rigorous, cautiously scoped audit study.
ActMap: Single-Pass Uncertainty Quantification from Generation-Time Activation Maps
Jacopo Dardini, Roberta Calegari
ActMap compresses generation-time hidden states into a fixed 96 KiB map; strong in-domain results, but near-chance transfer limits deployment claims.
IMM-based Multiple Object Tracking using a State Prediction Neural Network
Chan-Bin Lim, Dong-Hee Paek, Seung-Hyun Kong
KAIST's PR-IMM adds a Transformer predictor as a fourth IMM mode; strong RadarScenes gains, but oracle track IDs and no significance testing limit it.