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 One-Size-Fits-All: Sample-Adaptive Strategy Routing for Vision Token Pruning in MLLMs
Haiji Liang, Pengfei Zhou, Zhenglin Wan, Wei Wang et al.
Proposes per-sample pruning-strategy routing, +26.9% rel. accuracy over best fixed on VTC-Bench Group A; solid on its subset, real-world routing untested.
Cyber-Financial Contagion: Modeling the Propagation of an AI Vendor Compromise Through the Banking System
Alex Leytes
Timely synthetic model linking AI-vendor compromise to interbank contagion; reproducible, but simulator-derived labels and tiny vendor sample weaken claims.
ConvMem: Convolutional Memory for Long-Context Reasoning
Hongming Zhang, Zhaozhen Gu, Fengshuo Bai, Ming Hao et al.
Claims training-free hierarchical 'convolution' beats training-free baselines; fair setup, but latency and RL-overfitting claims outrun the shown evidence.
Why Is Video Still So Expensive? A Survey of Inference-Efficiency Mechanisms in Video and Audiovisual LLMs
Killian Steunou, Yannis Tevissen, Mounîm A. El Yacoubi
Survey of 125 works mapping VideoLLM efficiency by pipeline stage; disciplined comparisons, but evidence rests on borrowed re-runs, not new experiments.
Searching for New Physics with Reinforcement Learning
Jacky Kumar, Marianne Bouchard, David London
RL actor-critic autonomously searches 912 SMEFT operators to explain anomalies; reproduces W-mass results and beats random search, but no code release.
A Later Test Set Is Not a New Domain: Pretraining Familiarity Survives a Contamination-Free Hold-Out
Mahdi Naser Moghadasi, Faezeh Ghaderi
Claims pretraining familiarity survives a contamination-free hold-out; credible protocol and honest stats, but corpus cause rests on one model-domain match.
PACE: Perceived-Latency-Aware Cascading Service Routing and Filler Control for QoE-Efficient Retrieval-Augmented Dialogue Serving
Lin Huang, Yujuan Tan, Weisheng Li, Lixiang Zeng et al.
Claims a load-adaptive cascade with filler control halves P95 perceived latency in deployed RAG dialogue serving; credible system study on self-built data.
Understanding LoRA Rank Trade-offs in Diffusion Model Fine-Tuning
Iman Khazrak, Narges Nejad, Mostafa M. Rezaee, Robert C. Green
Claims moderate LoRA ranks (4/8) are best under fixed diffusion fine-tuning budgets; controlled but single-run, tiny-scale, with an unexplained anomaly.
Emergency Department Revisit Quality Review Screening: Exploring Human Decision-Making and Artificial Intelligence Support
Jonathan A. Handler, Marlene I. Robles-Granda, Jacob E. Mefford, Jeremy S. McGarvey et al.
Study finds GPT-4 over-flags ED revisits (94% yes) but a diagnosis-pair knowledge-graph screen shows 83-100% PPV; promising yet tiny, in-sample, unvalidated.
HybridFLow: SDN-Orchestrated Client Partitioning for Hybrid Federated Learning
Osama Abu Hamdan, Rabin Pandey, Hao Che, Engin Arslan et al.
SDN telemetry guides hybrid-FL sync/async partitioning; coherent closed-loop design, encouraging emulated gains, but key ablation and stats are missing.
Quantifying the Memorization-to-Generalization Transition: Scaling Laws and Phase Structure in Grokking
Anish Kataria
Empirical scaling law for grokking time in MLPs (D dominates H) is valuable, but internal inconsistencies and censored-data fitting weaken confidence.
Retrofitting Code Using LLMs to Support Exceptional Behavior
Linghan Zhong, Jiyang Zhang, Jayanth Srinivasa, Junyi Jessy Li et al.
Proposes retrofitting code with exception logic from tests; consistent large LLM gains across 5 models, yet ~25% of test-passing code differs from intent.
Multi-Agent Reinforcement Learning for Autonomous UAV Exploration in Wildfire Response
Caden Chandra, Jerry Ng
Sim-only multi-agent DRL shows fire-boundary tracking after 1000 episodes, but n=1 runs, no baselines, and missing hyperparameters cap confidence.
Rosetta at AlexandriaX-2026: LoRA-Adapted NileChat for Context-Aware Dialectal Arabic Dialogue Translation
Nada Esmaeil, Fathima Rena, Sibi Subhash, Osama Elgendy et al.
A shared-task system paper: LoRA-tuned NileChat-3B ranks 4th/5th; its honest negative-transfer finding is valuable but rests on untested point estimates.
Nonmaximal sums of maximally monotone operators under Rockafellar's constraint qualification
Weifeng Yang
Claims first disproof of Rockafellar's sum conjecture via explicit c0 and ℓ1 counterexamples; internally consistent, needs independent verification.
Algorithmic stability via ensembling
Rina Foygel Barber, Richard J. Samworth
Operator-norm theory shows ensembling by averaging stabilizes any algorithm to arbitrary data perturbations; clean and tight, but assumes bounded outputs.
Fortunate Recall: Ontology-Driven Memory Lifecycle Management for Persistent Coherence in LLMs
Ansuman Mullick, Eray Tüzün
Ontology-driven LLM memory lifecycle; pre-registered ablations credit generic metadata with correctness, ontology with calibration; LLM-judge-only validation.
Forgetting Only What Matters: Layer-Selective Unlearning toward Robust LLMs
Ravi Ranjan, Olivera Kotevska, Agoritsa Polyzou
Layer-selective LLM unlearning with claimed quantization robustness; favorable utility trade-offs, but 4-bit M1 ties SURE+NPO and reporting gaps weaken it.
Cross-Model Agreement as a Deployment-Time Reliability Signal for Automatic Polyp Segmentation
Siddharth Gupta, Jitin Singla
RBQE: independent-referee agreement flags polyp-seg failures label-free (AUC 0.876–0.960); strong ablations, yet one primary model and shared training data.
GEOSTEER: Geodesic Optimization for Activation Steering in Large Language Models
Xuan Cuong Ngo, Hao Vo, Ngan Le
GeoSteer: multistep geodesic norm-preserving steering via a learned probe; wins across 4 LLMs and 3 tasks, but no variance or significance tests reported.
JarvisGUI: Towards Cross-Device GUI Agents with Dynamic Task Composition
Zixiang Chen, Yuheng Lu, Zihao Cheng, Zeming Liu et al.
First cross-device GUI benchmark across 3 OSes with typed task auto-composition; near-zero multi-device scores may partly reflect the shared planner harness.
Do speech foundation models really learn words?
Robin Huo, Ewan Dunbar
Using ridge residualization of phoneme/diphone/triphone information, the paper argues HuBERT and wav2vec 2.0 encode words beyond phonetics; controlled and mostly convincing, though not fully airtight.
Building Multilingual Bridges: Data Mixing as the Pillar of Generalization for In-Language Reasoning
Mehrnaz Mofakhami, Ananya Sahu, Alejandro R. Salamanca, Daniel D'souza et al.
SFT data mixing gives a 3.35B model >93% in-language reasoning over 60 languages at small accuracy cost; solid ablations, but L2 rate needs a trained trigger.
Semigroup-JEPA: Latent Dynamics Consistency for Zero-Shot Physics Generalization
Andy Zeyi Liu, Haoran Sun, Lucas Baker, Randall Balestriero et al.
Credible zero-shot OOD physics generalization via rollout-trained JEPA; strong ablations, but single-seed evidence and baseline conflation temper claims.