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.
MeshHeal: Two-Timescale Self-Healing for Gray Failures in Decentralized LLM Agent Networks
Keru Chen, Sen Lin, Yingbin Liang, Nathaniel D. Bastian et al.
MeshHeal's two-timescale peer review heals gray LLM-agent failures; thorough evaluation, but injected degradation (120B→1B swap) is extreme, easing detection.
CrossSafe: Towards Cross-Embodiment Latent Safety Filters
Ihab Tabbara, Yuxuan Yang, Hussein Sibai
Proposes a shared HJ safety filter generalizing zero-shot across 5 robot embodiments; promising sim results but high residual collisions and muddled ablations.
Where Hallucinations Live: A Cross-Architecture Circuit in VQ-Tokenized Vision-Language Models
Shamanthak Hegde, Xiangrui Liu, Maitreya Patel, Yezhou Yang
Shared L0 routing circuit claimed to cause VQ-VLM hallucination; rigorous causal induction, but generative gains on one model and thresholds look post-hoc.
Shared Global KV with Layer-Specific Local History
Xinglang Xian
Reports 1.4% PPL gain from local history beside shared global KV at 126M, rigorously controlled; tiny scale and n=2 design limits transfer.
PISCES: Physics-Informed Solar-wind Convolutional autoEncoder for Space-weather Anomaly Detection and Early Warning
Kevin Lee, Alison J. March
PISCES (physics-informed conv AE for L1 solar-wind anomalies, OMNI 2018–24) reports AP 0.365–0.406 and 24–28-min warning leads, but a parameter-free energy baseline reaches 0.341 and no result is blind.
Evaluating Open-Weight LLMs for Turkish Domain Documents Under Retrieval and Hardware Constraints
Imtiaz Ul Hassan, Öykü Akbulut, Onur Kaya, Ardhendu Behera et al.
Careful benchmark of five 7-8B open-weight LLMs on Turkish document QA on a 6GB GPU; sound evidence-annotated protocol, but two documents limit reach.
How Much Were You Told? Measuring External Information in Peer Reviews
Matthieu Dubois, Pablo Piantanida, François Yvon
Self-Conditioning: unsupervised external-information score for peer reviews; separates delegated from polished reviews, but validated on synthetic regimes only.
Improving Ensemble Filters with Flow Matching
Haoyuan Chen, Alexandre Thiéry
FlowEF learns a flow-matching analysis step atop ensemble filters; strong theory and consistent gains on chaotic systems, but only linear observations tested.
SGA: Uncertainty Quantification for Multi-Step Forecasting in Time Series Foundation Models
Xin-Yu Hu, Shuang Liang, Cheng Feng, Shao-Qun Zhang
SGA quantifies TSFM forecast uncertainty via DAG complexity; broad 11-model × 27-dataset wins, but its scaling-law and near-random-baseline claims overreach.
TEMPS: Temporal Sentence Embeddings for Temporal Information Retrieval
Mourad Hassani, Julien Romero, Amel Bouzeghoub, Christian Jacquelinet
Proposes TTS task and TEMPS, a weakly-supervised Gaussian temporal branch for dense retrievers; gains are real on TimeQA/TS-Retriever but mixed on TempReason.
Tensor Decomposition of Transformer Key-Value Caches: Spectral Structure and Format Comparison
Rahul Krishnan, Volker Schulz
KV-cache head/layer modes resist compression, token/feature modes compress; Tucker wins at matched storage; solid spectra, but proxy metrics and single T=1024.
Prompt, Probe, Train, or Annotate? Single-camera sports video understanding in amateur settings
Sai Varun Kodathala, Prashanth Pollishetty, Jaylen Cargill
4-paradigm study on 66 amateur volleyball matches: rules beat perception, identity resists all methods; compelling but unverifiable (private data, no code).
A Native-Reference Coordinate Geometry for L2 Pronunciation Deviation Using Self-Supervised Speech Models
Tina Raissi, Nhan Phan, Mikko Kurimo
A training-free native-reference SVD geometry tracks L2 English proficiency (ρ≈-0.5); sound and honest, but tuned on train/dev only with no held-out test.
SlackDrive: Reclaiming Runtime Slack for Adaptive Driving Inference
Xiaohuan Pei, Hengguang Zhou, Yuanhao Ban, Justin Cui et al.
Latency-feedback compute allocation for driving models: simple, well-analyzed, but the 21.7% headline appears in no table and key details are missing.
Exact Quantile Balancing and Load-Error Injection for Mixture-of-Experts
Pit Neitemeier, Jiaze Li, Alessio Serra, Philipp Scholl et al.
EQB/LEI deliver exact global quantiles and local load-error gradients for MoE routing; 500B-token gains are promising but rest on one seed and one scale.
Reference-Based Analysis of Coherence and Diversity in Open-Ended Text Generation
Esteban Garcés Arias
Reference-based coherence/diversity evaluation via DTW alignment, summary gaps, likelihood; rigorous and honest, but no predictive gain over simpler baselines.
Curriculum Learning with GNN-based Reinforcement Learning for Job Shop Scheduling
Jayakrishnan K. Vasudevan, Jonathan Hoss, Noah Klarmann
Claims size-based curriculum cuts GNN-RL JSSP training time and lifts 30x30 quality; controlled, but 3 seeds and a weak heuristic baseline cap confidence.
LAYERSCOPE: A Layerwise Characterization of Video and Multimodal Learned Representations
Sandra Arcos-Holzinger, Debashish Chakraborty, Rohita Mocharla, Will Walden et al.
Label-free geometric audit of layerwise representations in 7 video/multimodal models; honest, useful, but narrow task coverage and weak statistical testing.
Discovery of fully efficient fault indicators along a data-based diagnosis process
Igor Bezmaternykh, Louise Travé-Massuyès, Elodie Chanthery
DT4X+ adds a fragmentation loss to ARR-inspired symbolic-regression diagnosis trees; dynamic-benchmark gains are real but small; text/table numbers mismatch.
Can LLMs Catch a Rigged Backtest? A Clean-Control Calibration Benchmark
Makar Ulesov, Vladislav Smirnov, Omar Ibrahim, Arsenii Bobovnikov
Paired clean-control benchmark: LLM backtest auditors saturate recall but clean-flag rates span 79 points; sound design, items unreleased, no human baseline.
Visual Tripwires: Anticipating Failure in Deep Vision Systems
Anoushka Harit, Rehan Zuberi, William Prew, Florian Markowetz
Uses latent temporal instability to predict vision-model failure early; promising idea, but synthetic sequences and weak temporal baselines limit the evidence.
Fed-ReMasker: Federated Tabular Imputation under Feature-Level Missingness
Ioannis Papathanail, Rooholla Poursoleymani, Lubnaa Abdur Rahman, Stavroula Georgia Mougiakakou
Applies ReMasker via FedAvg to federated feature-level imputation; rigorous MCAR benchmark shows clear wins, but novelty is modest and privacy untested.
Distillation for Efficient Multitask Manipulation Policies via Conditional Flow Matching
Shreya Deshmukh, Imen Mahdi, Nick Heppert, Abhinav Valada
Distills single-task CFM experts' velocity fields into one fixed-size multitask policy; promising RLBench gains but single-seed, simulation-only evidence.
Persistent Billable State: Denial-of-Wallet Attacks and Defenses in Tool-Calling LLM Agents
Jinqian Zhang, Haojun Xia, Shujiang Wu, Jingkun Yue et al.
Well-instrumented first study of DoW attacks on tool-calling LLM agents plus host-side D1–D4 defenses; credible but self-benchmarked, with no released artifacts (152 chars)