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.
Learning from VAE Errors to support ECG-based Differential Diagnosis of Myocardial Scar
Shayan Sharifi, Riccardo Treu, Ilaria Gandin, Federico Garoia et al.
Claims VAE reconstruction errors flag myocardial scar from ECG (AUROC 0.643): honest but modest results; small cohort, grid-picked configs, low specificity.
Optimal Rates for Agentic Networked Information Aggregation
MohammadHossein Bateni, Zahra Hadizadeh, MohammadTaghi Hajiaghayi, Mahdi JafariRaviz et al.
Claims and proves matching Θ(M²/D) rates for M-covered prediction paths, closing the KRR26 gap; visible proofs are careful — a solid theory contribution.
Online Change-point Detection for Cooperative Multi-Agent Reinforcement Learning
Fatemeh Saberi Khomami, Julita Vassileva
PPR (SMA→EMV→KSWIN on episodic returns) balances delay vs alarm stability in cooperative MARL, but 3 seeds and no external baselines weaken the evidence.
LexFlip: A Dissociation Diagnostic for Legal Meaning Preservation Metrics
Gaurab Baral
LexFlip's 373 French edits reversing legal force expose overlap metrics as blind; on FrJudge a length feature beats them. Unannotated labels; French-only.
How Does mHC Use Its Residual Streams? Selective Routing and Near-Identity Mixing
Pengxiang Zhao, Xing Li, Xianzhi Yu, Wei Guo et al.
Shows a 284B mHC model uses only part of its four-stream capacity; coherent diagnostics and interventions, but single-checkpoint scope limits generality.
The History Is the Detector: Executing CVE Patch History, End-to-End
Qiushi Wu, Kevin Eykholt, Youngja Park, Xiaokui Shu et al.
Claims CVE history becomes executable detectors; strong staged evidence, but headline 644 findings rest on partial validation and a 23-case patch eval.
What Matters, When? Diagnosing and Improving Conditional Visual Grounding in Visuomotor Imitation Policies
Vivek Chavan, Pengtao Xie, Yahuan Shi, Oliver Heimann et al.
Argues visuomotor distractor failures stem from phase-dependent grounding, not lost motor skill; interventions largely restore robustness; stats remain thin.
LLM-Driven Algorithm Design for Quantum Circuit Synthesis based on Binary Decision Diagrams
Yoonju Sim, Federico Berto, Chuanbo Hua, Jinkyoo Park et al.
Proposes QuantumEvo: LLM-evolved BDD variable-ordering heuristics targeting quantum cost; wins are real but modest, strongest off the search distribution.
Adaptive Gated Deepfake Detection for Low-Resolution and Resource-Constrained Environments
Vaishnavi Sen, Cody Laurie, Rashida Hasan
Quality-gated dual-branch multi-exit deepfake detector; Celeb-DF AUC 0.9370 beats 2 of 3 baselines, but gate routing stats unreported and FF++ near-chance.
Embedded Graph Flows for Categorical Graph Generation
Ethan Ma, Zihan Wang, Chris Siu Yeung Chow, Xinguo Feng et al.
EGF learns endpoint anchors for categorical graph generation, beating DiGress and GruM on all four QM9 metrics; yet the learned-geometry claim lacks ablation.
Technical Manual for a Toolkit for Measuring Contextual Individuation in Transformer Language Models
José Luciano Verçosa Marques, Frederico Jorge Heitmann, Daniel Omar Perez, Marcelo Vinicius de Paula et al.
Manual for a toolkit testing whether transformers individuate word senses by context via 'bridge forms'; design well argued, but lacks empirical validation.
Large Language Models for HVAC Operations in Building Energy Systems: A Critical Review of Methods, Applications, and Deployment Readiness
Alexander Neubauer, Tianzhen Hong, Han Li, Mengbo Yu et al.
Claims LLMs fit HVAC as semantic layers not controllers: 66 studies, 0 ready-now; credible review, but limited by single-coder Scopus-only corpus.
Lightweight Vision Transformer Compression for On-Device Plant Disease Detection in Resource-Constrained Agricultural Field Conditions
Mahadev Sunil Kumar, Bhavika Gondi, Desaisetty Venkata Satya Sai Swapnith, Gangireddy Rahul Jogi et al.
Claims 54.5× ViT compression to 6.01 MB at baseline accuracy for chilli disease detection; honest ablations, but internal inconsistencies weaken the evidence.
Design Docs Are All You Need: An AI-native Machine-Learning Performance Tool
Samuel Kushnir, Kimia Noorbakhsh, Kavya Sreedhar, Liqun Cheng et al.
SMART regenerates an ML perf-model library from design docs via agents; promising, rigorously motivated, but zero quantitative validation is presented.
Does Your Agent's Memory Survive a Model Upgrade? A Controlled Study of Memory Portability
Ankit Goyal, Jaideep Ray
Controlled study: KG-fixed memory survives model swaps, NOTES degrades sharply, mixed embeddings fail silently; rigorous but one 8B pair, synthetic histories.
Agents Trust Tools Too Much: Measuring Reliance on Unreliable Tools
Hoyeol Yang, Woojung Song, Taewon Kim, Jonghyun Song et al.
Finds 14 LLM agents adopt corrupted tool returns (68% for search) while hiding conflicts they detect; rigorous, but no intervention mitigates consistently.
Who Should Grade My Work? Student Perspectives on Transparent AI-Assisted Writing Assessment in Higher Education
Rayed AlGhamdi
Qualitative study of 13 Saudi students told ChatGPT graded their writing: feedback accepted, grading authority rejected; honest but small, single-site.
Distill Globally, Adapt Locally: Reasoning Distillation and Product-Type Test-Time Training for Scalable Trade-Up Recommendation
Siliang Liu, Mohammad Ghasemi, Sapan Patel, Amin Banitalebi-Dehkordi
LLM-distilled 15.5M classifier + per-type TTT reaches AUC 0.941 for trade-up; controls are honest, but pair-only splits and private data limit generalization.
WolfSociety: Understanding Collective Risk from Harmful-Agent Scaling in Financial Agent Societies
Lejun Zhang, Sarah Lu-Liang, Xin Jiang, Muning Wen et al.
Simulated agent societies: collapse boundary falls as αc∝N^−0.22; careful, honest study, but collapse is simulator-defined and the exponent is cutoff-sensitive.
RegionFed: Federated Learning for Personalized Query Understanding in Heterogeneous Retail Environments
Quoc H. Nguyen, Ali Lafzi, Abhijeet Phatak, Siddharth Pratap Singh et al.
Claims gradient-level FL personalization matches centralized accuracy on retail search; broad, well-ablated study but with internal inconsistencies.
Endogenous Exploration in Reinforcement Learning with Intrinsic Curiosity
Armando Vieira
LSM coherence model shows an interior 'curiosity window' for exploration; promising idea, but the claimed AIF comparison is absent and RL evidence is thin.
When LLM Decompilers Recompile More and Preserve Less
Chang Liu, Edward Raff, Kristopher Micinski
Shows passing shipped tests can mask behavioral divergence in LLM decompilers; rigorous fuzzing-based oracle, evidence broad but selection-biased.
Molecular Déjà Vu: Digit-Level Retrieval of Published Values in Frontier Language Models
Matthias Busch, Marius Tacke, Sviatlana V. Lamaka, Mikhail L. Zheludkevich et al.
Claims frontier LLMs verbatim-retrieve published molecular values digit-by-digit, amplified by reasoning; the audit's controls and statistics are rigorous.
Necessary or Sufficient? Evaluating LLM Explanations With Behavioural Evidence
Urja Pawar, Rajitha Ramanayake, Nabeel Kemal, Ashwin Kandath et al.
Asks if LLMs' cited top-3 features match measured necessity/sufficiency of their decisions; sound design, but synthetic data, 3-trial scores, no data release.