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.
FedGuide: Diffusion Prior Alignment and Value Baseline Guidance for Heterogeneous Federated Reinforcement Learning
Zhilin He, Gauri Joshi
FedGuide routes diffusion priors via OT-MoE with a DICE value baseline for heterogeneous FRL; promising results, but ablations and no significance tests temper claims.
From Foundation Embeddings to Cropland Maps: Label Efficiency, Temporal Transferability and Independent Human Validation
Mohammad Ammar Mughees, Giovanni Montefoschi, Zhongxin Chen, Maria Antonia Brovelli
A carefully scoped, statistically sound empirical study showing frozen AlphaEarth embeddings plus a lightweight classifier map cultivated land in Maine at ~93.7% OA and outperform the CDL against blind human validation, with honest limits but only single-region evidence.
Temporally Consistent Graph Extraction and Matching for Longitudinal Angiographic Images
Linus Kreitner, Laurin Lux, Carmen Baumann, Daniel Rueckert et al.
Sound match-before-refine idea for longitudinal vessel graphs, carefully engineered; evidence stays qualitative on private same-day healthy-eye OCTA data.
G3AR: Graph-Guided Neural Visual Geometry for Scalable Multi-Sequence Aerial Registration
Jeng Wen Joshua Lean, Ting-Yu Yen, Wei-Fang Sun, Simon See et al.
Claims scalable graph-guided neural geometry for aerial registration; matched-budget wins over VGGT-Long/MERG3R are credible, but accuracy still trails COLMAP ~20x.
HLC-GS: Risk-Map-Guided Height-Layer Consistency Gaussian Splatting for DSM Reconstruction from Optical Satellite Imagery
Jie Yang, Yingdong Pi, Qiyan Luo, Xiaoyu Wang et al.
Extends EOGS with risk-guided height-layer consistency losses for satellite DSM; good gains on 7 AOIs, but no significance tests, many fixed thresholds.
A Lightweight CNN Integrated Compact Convolutional Transformer for Multi-Scale Feature Learning and reducing computational complexity for breast cancer mammography image detection and classification
Md Taimur Ahad (Department of Management North South University, Dhaka, Bangladesh), Ainuddin Ahmed (Department of Management North South University et al.
Proposes a 250k-parameter CNN-CCT hybrid claiming 99-100% mammography accuracy; source-label 5-class task, leakage-prone splits and no baselines undercut it.
Making AI-Assisted Claims Independently Challengeable: Publication Authority and a Protocol for Falsifiable Publication Records
Torsten Olivi Tiltack, Yifei Dong, Kun Yu, Xu Wang et al.
Claims a fail-closed, single-use publication-authority protocol (PAC-2026 SF-4); rigorously bounded and honest, but evaluated solely within the authors' own lineage.
Too Good to Be Real? Diagnosing and Reducing the Gap Between AI Preference and Real User Engagement
Xinglang Zhang, Yuanmeng Xiang, Yunyao Zhang, Zeliang Chen et al.
Claims LLMs overweight explicit logic vs real engagement; OMRA cuts the gap 54.4% — coherent diagnosis, but measured in the authors' own ontology space.
From Pixels to Pairs: A Comprehensive Benchmark of LLM-Based Key-Value Extraction in Noisy Document Settings
Zahra Anvari, Vassilis Athitsos
一个有用的 LLM 键值提取在 OCR 噪声下的受控基准；发现合理，但关键匹配的不透明性和内部异常削弱了结论。
PANORAMA: Panoptic Grounded Captioning via Mask Proposal Selection
Sara Pieri, Evangelos Kazakos, Shizhe Chen, Josef Sivic et al.
Introduces PanoCaps benchmark and proposal-selection grounding model; strong, well-ablated results, but the headline gPQ margin sits near seed noise.
Decodable but Misrouted: Sparse Features Uncover a Readout Gap in Vision-Language Models for Harmful Meme Detection
Girish A. Koushik, Diptesh Kanojia, Helen Treharne
Claims sparse-autoencoder probes decode harmful-meme evidence that Gemma-3/Qwen3.5 fail to route to outputs; unusually candid and carefully scoped, but the headline gap pits a frozen zero-shot model against supervised in-distribution probes, and causal evidence rests on 5–15-example interventions.
Visual Input and Its Framing Affect Attribute-based Descriptions Produced by Large Vision-Language Models
Xiaomeng Wang, Martha Larson, Zhengyu Zhao
Visual input shifts LVLM attribute descriptions toward physical terms (18% to 45%) despite correct recognition; a carefully controlled study but only 2 models and 30 pet breeds.
DiT-Garment: Garment Dynamics with Diffusion Transformers
Antoine Dumoulin, Laurence Boissieux, Joao Regateiro, Pierre Hellier et al.
Diffusion-transformer garment animation generalizes to unseen designs; shape results beat DPF/ContourCraft, but 0.4 FPS contradicts the interactivity claim.
PRISM: Predictive Representation of Interaction Style and Motion for Social Robot Navigation
Bo-Han Chen, Hiromu Taketsugu, Norimichi Ukita
RnC-based passive style inference cuts crowd-sim collision rates vs HEIGHT in all regimes; promising, but simulation-only with no CIs or seeds.
RecMorph: Topology-Guided Spatial Recurrence for Generalized Morphology Control
Quanrui Rao, Yong Liu, Xueming Xiao, Yingbo Luo et al.
RecMorph: topology-guided bidirectional recurrence shows strong UNIMAL and sim-to-real gains, but evidence rests on 4-seed means with no significance tests.
BLADE: ReliaBle Dynamic Hardware-Aware SNN-ANN Boundary SeLection for Event-BAseD Object DEtection
Mahdi Taheri, Alwin Paul
First reliability-aware SNN-ANN boundary/early-exit co-selection; rigorous fault study with honest nulls, but analytic energy and single-dataset validation.
ORCA: Occlusion-Aware Refinement and Completion for Novel View Synthesis
Weronika Jakubowska, Maciej Zi\k{e}ba, Przemys{\l}aw Spurek
ORCA repairs novel-view holes with local RGB-D or generative inpainting by size, beating VistaDream on all metrics; sound method, thin evaluation.
Colla-Q: Toward Collaborative Experts in MoE Quantization via Minimax Precision Balancing
Eunju Shin, Jongbin Ryu
Balances expert performance in MoE quantization via an activation-entropy ratio and greedy minimax bit allocation; well-ablated with strong calibration robustness, but gains over PMQ are often under one point.
Semantic-Spatial Agreement Verification for Mitigating Object Hallucination in Multimodal Large Language Models
Ziheng Ren, Qian Gao, Jun Fan, Guohui Ding et al.
SSAV checks MLLM object claims via multi-prompt, region-consistent detector evidence; solid ablations, but calibration and per-model tuning details are thin.
Walking the Score Manifold: Continuous-time Generative Dynamics on Learned Data Manifolds
Jan Tauberschmidt, Brian B. Moser, Stanislav Frolov, Andreas Dengel et al.
Score-geometry-guided continuous-time generation with a principled contraction objective; strong 2D/PDE/MD evidence, mixed KTH video results, single-run evals.
Tracing individual knowledge trajectories in a changing field: the case of general relativity and gravitation
Raphael Schlattmann, Malte Vogl
A rigorous, reproducible 50-author GRG study finding vocabulary tracks later field states and semantic density earlier ones; headline counts lack null models.
Monitoring and Discovering Reward Hacking with Internal Representations during LLM Evaluations
Leon Bergen, Usha Bhalla, Andrew Lee, Barak Widawsky et al.
Synthetic-data DoM probes detect reward hacking in frontier LLMs nearly as well as LLM monitors at near-zero cost; solid work resting on an LLM-judge ground truth.
ProgramDistill: From Interactive Web Apps to Verifiable Reference-Guided SWE Tasks
Jeonghye Kim, Minseon Kim, Young Jin Kim, Matheus Pereira et al.
Automated pipeline turns 26 web apps into 4,063 replay-verified agent tasks; strong engineering, but thin statistics and no public release yet.
Hypothesis-Driven Autonomous Materials Synthesis with Multimodal LLM Agents
Izumi Takahara, Kazunori Nishio, Akira Aiba, Shigeru Kobayashi et al.
SynAgent lets LLM agents run a physical synthesis lab and produce testable understanding; convincing single-campaign demo, but no BO baseline, no replicates, prior-leakage risk.