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.
RoboSPA: Can VLA Models Go Beyond Simple Scenes and Short-Horizon Tasks?
Zhenxuan Fan, Bo Zhang, Yutong Lin, Yuqian Yuan et al.
Claims a 527K-trajectory, 280-variant benchmark diagnosing VLA spatial/planning failures; rigorous and diagnostic, but experiments use only a slice of it.
Evaluation of MLLM-Agnostic Plug-and-Play Keyframe Selection Methods for Long Video Understanding
Dilip Sarkar, Md. Safayet Islam, Liang Liang
First head-to-head of five PaP keyframe selectors; authors' own QAaF wins 13/15, but an unpublished core method and encoder/budget confounds weaken it.
Diffs vs. Whole Files: An Empirical Comparison of Iterative Edit-Based and Direct Generation for Flutter/Dart Code Models
Andrej Andrejev
Claims direct whole-file generation beats diff-based editing on Flutter/Dart at two model scales; credible, though the qwen-direct LR bug and token asymmetry temper it.
SurveyAgent-HKA: A multi-agent framework for scientific survey generation with LLMs and human knowledge augmentation
Tong Bao, Mir Tafseer Nayeem, Yi Zhao, Davood Rafiei et al.
Multi-agent survey generation with human/peer-review knowledge tops baselines on citation and structure, but absolute citation F1 (~17–19%) remains low.
STAR-Pro: Stage-Wise Token Adaptive Reduction with Progressive Refinement for Efficient Large Vision-Language Models
Yichen Guo, Tinghao Wang, Qizhe Zhang, Lingbei Meng et al.
STAR-Pro: training-free two-stage visual token pruning for LVLMs with strong wide-scope results; wins mostly hold up, but margins lack significance testing.
Deep Microcompression: Structured Pruning and Bit-packed Quantization for Microcontrollers
Opegbemi Matthias Busoye, Tolulope Matthew Busoye, Eghonghon-aye Eigbe
Claims 55.8x-compressed LeNet-5 on a 2KB-SRAM MCU via pruning+QAT+bit-packing; credible engineering, but single-benchmark eval and 'first' claim need scrutiny.
An Evolutionary Computation Framework for Multi-Agent Q-Learning with Mean-Field Environmental Feedback
Lichen Wang, Shijia Hua, Linjie Liu
Derives a mean-field transport equation for graph Q-learning coupled to an evolving environment; clean derivation, MC-validated, but tested model is stylized.
AgentBrew: Offline Tool-Use Agent Learning from Raw Real-World Trajectories
Zhiyi Lyu, Yewen Li, Longtao Zheng, Shengtian Yang et al.
Offline hindsight relabeling + PMI action credits lift Qwen3-32B by +8.7 Acc avg on 3 MCP apps; promising, but small evals lack variance reporting.
Accelerating Diffusion Transformers with Gaussian Process Rectified Feature Cache
Zhirong Shen, Rui Huang, Chang Zou, Shikang Zheng et al.
GPR-based plug-in corrects DiT feature-cache drift with uncertainty-gated recomputation; gains look strong, but abstract headline numbers don't match tables.
Tri-PvP: Exposing Modality Bias in Omni-Modal Large Language Models through Perceptual-Propositional Evidence Conflicts
Yen-Ting Piao, Shu-Yun Chen, Chin-Hui Chu, Chun-Wei Chen et al.
Solid tri-modal benchmark disentangling modality from evidence-form bias; credible results, but propositional evidence is confounded with TTS/rendering.
What Does Animal Re-Identification Learn? Linear Biological Concepts and Their Origins in Visual Representations
Robert Nolting, Alexandra Schild, Moritz Weckbecker, Maximilian Schall et al.
Claims a gorilla Re-ID ViT linearly encodes sex/age (0.91 AUROC) and causally uses sex; credible probing+steering evidence, single-species, no CIs.
VERPO: Verified Evidence Regularized Policy Optimization
Haijiang Li, Chengyu Lv, Yi Zhang, Rui Qian et al.
VERPO adds Fisher-gated token-level evidence corrections to GRPO; gains are solid on 1B–8B models but thin on Qwen3-4B and come at ~1.7x compute.
Spectral Prioritized Sweeping in Nonstationary Reinforcement Learning
Hung Pham, Tuan Dam
GTA-PS adds a Laplacian-resolvent topology term to prioritized sweeping for nonstationary rewards; sound theory, tabular gains, single baseline only.
TruthInsightBench: An Evidence-Grounded Benchmark for Automated Evaluation of Open-Ended Scientific Discovery Agents
Zhibo Yang, Chen Zhang, Yuewei Zhang, Hao Wang
Discovery-oriented 40-task benchmark scored by an automated LLM-judge rubric; four agents on one base model plateau at 58-60/100; judge remains unvalidated.
NEAT-POCKET: Pocket-Conditioned Autoregressive 3D Molecular Generation with a Neighborhood-Guided Set Transformer
Roxane Axel Jacob, Daniel Rose, Thierry Langer, Johannes Kirchmair
Autoregressive pocket-conditioned 3D generator is fast and benchmark-competitive; credible, though docking scores lag and size confounds metric comparisons.
Measuring AI Accountability Through Argumentation Analysis: Can Model Reasoning Withstand Scrutiny?
Daan R. Henselmans, Derck W. E. Prinzhorn, Arno Libert
Novel ground-truth-free test of LLM moral defensibility via Walton schemes; solid validation, but judge self-overlap and uncorrected tests temper findings.
ProCA: Progressive Contrastive Alignment for Robust EEG Visual Decoding
Kanglei Zhou, Chunyan Lan, Dongyang Li, Jun Zhu et al.
Progressive confusion-guided alignment + structured mixup for robust EEG decoding; gains consistent across protocols but relative numbers mask small absolutes.
LLM-Guided Program Evolution for Circle Packing: Breaking 10 Packomania Records for $28
Wes Sander
Claims a $28 LLM evolution loop broke 10 packing records, but Table 1 shows its classical seed solver broke all 10 at iter 0; LLM gains were only ~1% overall.
NS-ST-GraphRAG: Neuro-Symbolic Spatio-Temporal GraphRAG for Literary Knowledge Processing
Zheng Kui Lin
Temporal GraphRAG + claimed-first QA benchmark for a Chinese classic; honest but key gain 0.733 vs 0.675 (p=0.092 n.s.); GraphRAG baselines deferred.
H2LooP Telecom Model v1: From Telecom Comprehension to Autonomous Issue and PR Resolution
Amit Singh, Vedant Nipane, Mayank Goel, Pulkit Agrawal et al.
31B telecom SFT claims OT-Lite 81.8% and telco-leaderboard #5, ahead of frontier LLMs; plausible but weakened by unnamed base, ±5pp noise, private eval.
Unifying ICL, SFT, KL-Regularized RL Through a Bayesian Lens
Junxin Fan
Paper unifies ICL, SFT, KL-regularized RL as one Bayesian template: Gibbs posterior + forward-KL projection; sound synthesis of known results, small tests.
Single-Query Black-Box Calibration Auditing via Logit Bias
Roman Plaud, Antoine Saillenfest, Matthieu Labeau, Thomas Bonald et al.
Clever one-query calibration audit of black-box LLMs via logit_bias thresholds; theory sound, simulated results strong, but no live-API test and no error bars.
Machine learning-assisted calibration of Agent-based Models: surrogate-based optimization with Genetic Algorithm and Particle Swarm Optimization
Duguma Yeshitla Habtemariam, Jihwan Lee
Careful 48-config transfer of inner-loop surrogate-assisted GA/PSO to ABM calibration; speed gains robust, accuracy gains best-case only, honestly reported.
A Comparative Study of Counterfactual Explainers for Graph Neural Networks Enabling Multiple Types of Graph Edit
Maria Myrto Villia, Filippos Gouidis, Theodore Patkos, Panos Trahanias
Benchmarks six counterfactual explainers for GNNs incl. add+remove methods; useful first head-to-head, but single-run stats and table anomalies weaken rigor.