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.
EMR: Self-Evolving Medical Multi-Agent System via Experience Mining and Reuse
Dongsheng Shi, Yue Li, Xin Yi, Linlin Wang
Claims medical multi-agent self-evolution via experience mining (81.1 vs 77.8 avg acc, Qwen3-8B); strong ablations but no significance tests or code.
PACE: Progressive Angular-to-Norm Contrastive Embedding
Yanping Li, Wei Zhou, Yawen Liu, Yibo Wang et al.
Claims progressive cosine→dot and LoRA→full training stabilizes dot-product multimodal embedding; +2.0/+2.4 MMEB gains plausible but no significance tests.
Improving Mathematical Reasoning Capabilities in Large Language Models via Reasoning Process Error Classification
Runa Yoshida, Kosuke Nishida, Kyosuke Nishida
Manual 21-class taxonomy of MATH reasoning errors plus an error-focused prompt yields significant gains across 3 LLMs; sound analysis, single-benchmark scope.
Temporal Self-Distillation: Faster Inference in Discrete Diffusion Language Models
Shijian Xu, Andrea Miele, Metod Jazbec, Volker Roth et al.
TSD, single-stage on-policy self-distillation across denoising steps, shifts dLLMs' speed-quality frontier to low NFE on 7 benchmarks; results lack error bars.
Multi-source Transfer Learning of Time Series with a Shapelet-based Distance Measure
Jiseok Lee, Brian Kenji Iwana
Proposes training-free shapelet-based source selection for multi-source transfer in TSC; rigorous UCR study, but its edge over DBA-DTW is not significant.
STHMoE: Hypergraph-Enhanced Heterogeneous Dependency Coordination for LLM-Based Urban Traffic Data Forecasting
Jiawen Chen, Qi Shao, Yongjian Chang, Mingtong Zhou et al.
Claims SOTA via LLM experts + adaptive hypergraphs + soft MoE routing; competitive on 10 traffic benchmarks, but margins often small and no variance reported.
Impute-EM: Native Mixed-State Diffusion Models for Heterogeneous Data Imputation
Sergei Kholkin, Kirill Sokolov, Dmitry Baranchuk, Evgeny Burnaev et al.
EM-style imputation with native mixed-state diffusion backbones; strong tabular fidelity gains but incremental over DiffEM/DiffPuter and no error bars.
CITECHOICE: A Causal Audit of How Document Presentation Redistributes Citation Credit in Agentic Search
Sriram Selvam, Anneswa Ghosh
Hash-verified replay: structured rendering redistributes citation credit (+0.50, Holm p=.033); primary incidence null — rigorous, honest, scope-limited.
Nearly Minimax-Optimal Regret for Linear Contextual Bandits with Arbitrary Adaptive Action Sets
Tianyuan Jin
Resolves minimax K-dependence in adaptive linear contextual bandits with matching bounds; results solid but algorithms are exponential-time, not practical.
MAST: Label-Efficient, Robust, and Generalizable Sound Detection for Biodiversity Monitoring via Masked Audio Pretraining and Self-Training
Tianyi Xu, Daniel Pimentel-Alarcón, Zuzana Buřivalová, Claudia Solís-Lemus
MAST claims robust low-label sound detection via masked pretraining+self-training; gains are large but single-run, no CIs, baseline-collapse confounds.
What Limits Us? Analyzing Self-Reported Limitations in NLP Research
Tawan Thaepprasit, Peeranuth Kehasukcharoen, Ding Wang, Remi Denton et al.
Human-AI coding of 16,067 ACL/EMNLP Limitations sections (2020–25) finds scope-limitations surging; rigorous and honest, but validation rests on 150 papers.
Convergence rates for generative drifting flows: fixed-scale obstructions and multihead acceleration
Arthur Stéphanovitch, Eddie Aamari
Proves fixed-bandwidth drifting is intrinsically slow; multihead scale-averaging restores exponential convergence, though only locally near reference laws.
Rethinking Correctness for Uncertainty Estimation in Clinical Prediction with Vision-Language Models
Mingcheng Zhu, Jinning Liang, Tingting Zhu
Claims correctness criteria must be validated before UE comparison; two-axis audit favors EM over LLM-judge, but the 450-case audit limits generality.
MUSE: A Theory-Harnessed Story Engine for Vibe Narrativizing
Jianxiang Ma, Xiaocui Yang, Daling Wang, Yuesong Hou et al.
McKee-theory agent engine lifts LLM story quality on 2 of 3 benchmarks across 4 models; promising, but all-LLM judging and weak baselines temper it.
VisInteract: Towards Dynamic Interactive Text-to-Visualization under Imperfect Queries
Wenxin Xu, Jinwei Lu, Hwanhee Kim, Chen Jason Zhang et al.
Claims first interactive Text-to-Vis benchmark + MCTS method; large gains (+13.4–16.3pp) and solid ablations, but judge/user-agent circularity needs scrutiny.
Semiotic Relations and Proof Methods: A Cross-Genre Study of Argument Structure with Large Language Models
Edirlei Soares de Lima, Marco A. Casanova, Antonio L. Furtado
Paper maps four master tropes to four proof methods and finds uneven cross-genre use via LLM annotation; intriguing but labels lack human validation.
Failure-Guided Co-Evolution of Prompts and Training Data
Tianyu Yuan, Zhuzhong Qian
FORGE co-evolves prompts with failure-guided synthetic data, +16.52 pts over baseline on 8 benchmarks; promising but single-seed, no significance testing.
Interpreting hierarchical organisation of speaker embeddings
Yanze Xu, Wenwu Wang, Mark D. Plumbley
Proposes HCCM and L-score to interpret SLINK hierarchies of speaker embeddings; interesting findings, but single model/dataset and no clustering baselines.
TEAR: Table Extraction with Attribute Recommendation from Texts via Large Language Models
Tong Li, Shuye Ding, Jiachuan Wang, Yongqi Zhang et al.
LLM framework TEAR adds text-driven attribute recommendation to table extraction on natural texts; solid benchmark and gains, but significance testing is thin.
Issue Bias in Generative AI Writing Assistance: Political Issues and LLMs in the Swedish 2026 Election
Bastiaan Bruinsma, Annika Fredén, Paul Röttger, Moa Johansson et al.
Six LLMs show issue-level stances in Swedish writing tasks—GPT neutral, Grok divergent—yet no significant party alignment; LLM-judge validation unfinished.
Pre-PEFT Probing: Weight Statistics and Perturbation Robustness for Layer Selection in VLM Vision Encoders
Qingtao Xia, Jiahua Bao, Siyao Cheng, Jie Liu
Weight stats and perturbation robustness signal which VLM encoder layers to adapt via PEFT; evidence is broad but the core correlation is never quantified.
Augmenting Large Audio-Language Models with Frame-Level Grounding for Fine-Grained Temporal Perception
Yanfeng Shi, Yan Song, Junhui Li, Tinggan Huang et al.
Freezes the LALM to encode queries and adds a frame-level grounding head; strong gains on 5 grounding benchmarks, but baseline fairness and stats are thin.
From Ideas to Actions: A Public-Data Decision-Support Toolchain Across the Venture Lifecycle
Lei Qu
Claims an auditable public-data founder-decision toolchain (F0.5 0.54 clean validation); transparent, but scale validation incomplete, Part Two exploratory.
Bandits with Probing: Optimal Regret and the Limits of Winner Feedback
Yongjie Guan
Resolves open probing-bandit problems with two tight minimax regret laws; coherent novel construction, though full proofs exceed what this text can verify.