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.
The Missing Complement: State-Conditioned Minimal Sufficient Evidence for Coding Agents
Zhexi Feng, Ruiyi Zhang, Yongbo Yang, Pengtao Xie
Formulates set-level sufficient-evidence retrieval for coding agents with a careful benchmark; main gains robust, downstream gains partly within noise.
FAMOS: Feed-Forward 3D Articulation Modeling from Sparse Observations
Kevin Qu, Tao Sun, Massimiliano Viola, Liyuan Zhu et al.
FAMOS reports strong feed-forward articulation estimation from sparse multi-state point clouds; gains are large, but key baselines are author-adapted.
Math for AI safety: an invitation for mathematicians
Lionel Levine
A survey inviting mathematicians into AI safety by field, with eight well-posed open problems; excellent curation, but no new theorems or evidence of its own.
Reason What Matters: Retrieval-Grounded Reasoning for Universal Multimodal Embeddings
Mingzhou Jiang, Peixi Wu, Hang Cheng, Yunhao Zhou et al.
ReWAM adds token-level retrieval credit (RASD) and adaptive CoT truncation (RAI) to multimodal embeddings; SOTA gains look real but thin and unseeded.
The Universe of Universes: Benefit Yield Functions, Implosion Thresholds, and Infrastructure-Aware Optimization in Multi-LLM Systems
Danielle Franklin, Vasu Raj Jain
Formalizes LLM-ensemble size effects (BYF, implosion threshold θ*); all evidence is synthetic simulation only, and the headline theorem is a proof sketch.
Admissable: Training Reinforcement Learning Agents against Adversarial Missingness
Paul Stahlhofen, Luca Hermes, Tim Kochs, Markus Vieth et al.
Formalizes adversarial feature masking in RL and trains SAC against it; promising but thin evidence: 3 MuJoCo tasks, 1 baseline, 5 seeds, no sig. tests.
When Agents Slow Down: Understanding LLM Agents' Test-Time Strategies via Elo-per-token Analysis
Kaiyuan Liu, Qiuyang Mang, Bo Peng, Wenhao Chai et al.
Elo-per-token framework measures agent test-time scaling vs sampling with clean theory and huge evidence, but human superlinearity may reflect selection bias.
Reducing the Output-Mode Gap in Speech Language Models via Joint-Output On-Policy Distillation
Daxin Tan, Dehua Tao, Chengxi Deng, Hanlin Zhang et al.
Defines output-mode gap in speech LLMs and cuts it via self-distillation; strong two-model results, but spoken gains are smaller and untested statistically.
Evaluation Metrics for Safe Reinforcement Learning
Lindsay Spoor, Aske Plaat, Thomas Moerland
Proposes violation-rate, normalized-cost metrics, tiers and CDF reporting for safe RL; strong 1,440-run study, but narrow scope and brittle tier thresholds.
Clean Scores, Buried Evidence, and Confident Wrong: A Receipt-Based Audit of Frontier Agentic QA
Luis M. Sánchez
Buried-evidence audit of six frontier agents: accuracy drops, cost rises; rare statistical candor, but n=41 and grader sensitivity cap the claims.
Concept-Grounded Reasoning with Prompt-Driven Localization for Interpretable Structured Report Generation
Xinyue Xu, Hongbin Lin, Juangui Xu, Hualiang Wang et al.
CORAL adds concept bottlenecks plus SAM3 masks to an MLLM for medical reports; BUS/X-ray gains promising but templated reports inflate BLEU; ablations thin.
Parameter-Efficient Adaptation of Pretrained Language Models for Time-Series Forecasting
Tamanna Kumavat, Georg Brunner, Kyriakos Flouris
Frozen GPT-2 patch-embedding forecasting matches specialized models' MASE range per systematic ablations; single-seed and ranks behind DLinear/PatchTST.
Approximating Smooth Functionals with ReLU Networks
Shuhao Jiao
Claims first nearly optimal ReLU rates for smooth functionals on Hilbert space via joint coordinate decay; theory looks coherent, experiments thin.
MAPS: Memory-Aware Predictive Scheduling Framework for Large Language Model Serving
Tiancheng Zhang, Yulin Chen, Yunfeng Zhao, Shaoyuan Huang et al.
Claims 42.6% mean/84.8% tail latency cuts for disaggregated LLM serving via device prediction + conformal calibration; promising, but only 2 decoders tested.
Representing Clinical Conditions on Vital Signs from Healthy Individuals using Latent Modeling
Rafael Pina, Varuna De Silva, Mindula Illeperuma
CVAE + latent classifier reshapes healthy vital signs toward ICU conditions; interesting ablations, but the gameable mean-distance metric undercuts the claims.
SlopShape: Identifying AI-Generated Commercial Web Content
Jochen Madler
Claims structure alone detects AI blog posts at 98.0 F1, robust to rewording; rigorous replication, but year and brief confounds limit interpretation.
WebCraftBench: Evaluating Web Application Generation from a Software Testing Perspective
Chenxu Liu, Zilu Zou, Peizhong Gao, Jiawen Tao et al.
Coverage-guided interactive benchmark for LLM web-app generation; well-validated (85.3% human agreement) though dataset and code remain closed-source.
Distilling Foundation Models for Agentic What-If Reasoning:Cost, Latency, and Governance in a Hybrid LLM+SLM Architecture
Sourish Dey, Aditya Kumar
Distills TabPFN to an 8.5K-param MLP in a hybrid LLM agent; latency 3.8-4.8x, attribution honest, but single-seed runs and confounded splits cap the evidence.
Divide, Consult, Conquer: Capability Laundering Through Aligned LLMs
Mark Russinovich, Blake Bullwinkel, Giorgio Severi, Cristian Ovadiuc et al.
Shows weak unaligned LLMs launder frontier capabilities via task decomposition; solid uplift design, but bespoke policy prompt and judge bias temper claims.
CodeTS: Verifiable Text-to-Time Series Generation via Executable Code
Xudong Yuan, Shunyu Liu, Tongya Zheng, Huiping Zhuang et al.
CodeTS turns Text-to-TS into Text-to-Code-to-TS with SFT+RLVR; benchmark wins, but auto-derived descriptions embed target statistics, favoring the method.
MUUNRiver-Bench: Diagnosing Relation-Dependent Music Retrieval with Multimodal Instructions
Zhancheng Guo, Congren Dai, Shangda Wu, Jianhuai Hu et al.
Claims first fused-modal text+audio music-retrieval benchmark; well-QC'd synthetic corpus, honest negatives, but rankings lack statistical support.
SkillLift: Learning Dense Rubrics from Sparse Oracles for Efficient Skill Evolution
Haoxiang Kang, Ming Wen
Claims cheaper skill evolution via a rank-aligned rubric surrogate; broad gains on 147 tasks, but 3-seed stats and reduced-scale baselines warrant scrutiny.
Graph Matching Relaxations and Amortization for Supervised Graph Prediction
Federico Méndez, Paul Krzakala, Gabriel Melo, Charlotte Laclau et al.
Amortized GW matching for graph prediction: tightness theory plus large gains; solid, but single-run results and uneven baselines limit rigor.
When Tool Calls Succeed but Workflows Fail: Anomalies at the Agent-Tool Boundary
Artem Trofimov, Boris Novikov
Defines 8 agent-tool effect anomalies; 98k-tool MCP census shows no annotation can express the needed transactional contracts. Strong taxonomy, proofs pending.