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 Similarity: Coverage-Aware Prompt Selection for Time Series Forecasting with LLMs
Daeun Ji, Minkyoung Kim, Dongkuk Kim, Yohan Lee et al.
Parameter-free coverage regularizer for LLM prompt retrieval in forecasting; unusually honest stats, but the aggregate gain resolves on only 1 of 24 cells.
Dual-Locking Learned AI Models: A PIN-Based Sparse QIM Watermarking and Adaptive Index Permutation Approach
Iva Vasic, Jesús Muñoz-Cádiz, Bata Vasic
Claims reversible weight-permutation locking + QIM bias watermarking secure AI models; promising results, but evaluation and security analysis lack rigor.
Connectivity-Aware Exploration of Robotic Grasp Spaces
Maksim A Kazanskii
Credible simulation-only study: connectivity-aware grasp-set sampling speeds structure reconstruction (2-5x) but adds little over local viability scoring.
LPINNs: First-Layer Gated Localization for Physics-Informed Neural Networks
Lakshay Chawla, Hardik Jain
PINN first-layer gating cuts error 1-2 orders on three toy benchmarks; clean paired-seed evidence, but post-hoc config choice and only a vanilla baseline.
Interpretable Multi-Hypersphere Deep Anomaly Detection for Open-set Supervised Anomaly Detection
Zhiji Yang, Fangyong Wang, Yue Li, Xianli Pan et al.
Per-class hyperspheres embedded in the final layer give interpretable open-set anomaly detection; strong AUC wins, but baselines are dated and single-seed.
Rethinking Pivot Programming Languages in Code Language Models
Andor Diera, Lukas Galke Poech, Matthias Tichy
Claims no universal pivot PL: geometry+retrieval show relation-dependent structure (Python=English-facing); well-controlled, modest scale.
OptiSkill: A Hierarchical and Evolving SkillBank for LLM-Based Optimization Modeling
Ruiqing Zhao, Rui Liu, Yuan Zuo, Huarong Zhang et al.
Solver-verified hierarchical SkillBank tops OR-modeling agentic baselines (73.9% avg, DeepSeek-V4); convincing, but no significance tests and only 2 backbones.
On attention heads and bilinear forms
Andrew O'Desky
Proves attention-head QK symmetric parts are balanced (n,n,N−2n) via a new profile map; theorems sound, but learned spectral patterns stay unexplained.
Reconstructed holograms and explanation-aware evaluation for low-cost computational pollen analysis in veterinary cytology
Swarn Warshaneyan, Joial Danyal, Blaž Cugmas, Mindaugas Tamošiūnas et al.
Hologram reconstruction cuts the brightfield-DIHM pollen-analysis gap substantially; XAI diagnosis is credible but rests on six test images and private data.
Auditing Political Alignment in LLM Assistants: Engagement, Stance, and User Identity
Joan C. Timoneda
Claims LLM political speech follows user-conditional 'speech regimes'; a preregistered 7,500-conversation audit is persuasive but rests on LLM judges.
Silent Failures Beyond the 32-Bit Index Range: A Differential Characterization of Large-Tensor Matrix Multiplication in PyTorch's MPS Backend
Junichiro Niimi
Rigorous black-box study finding silent MPS bmm corruption above 2^32 elements in PyTorch across 11 releases; systematic, reproducible, convincing.
Spatial-Interactor: Learning Spatial Reasoning through Interaction with the Observable Physical World
Kaixiang Yao, Xu Wang, Miao Pan, Hu Xiyue et al.
Interaction-trajectory training (LSI-108K; SFT+OPD) lifts VLM spatial reasoning broadly, but its own ablation shows generic spatial-QA SFT drives most gains.
Watching Quantum Models Think: Hilbert-Space Interpretability in Quantum Transformer Blocks
Diego Iacopetta, Andrea Gasparini
Proposes mutual-information tracking as intrinsic interpretability for a 12-qubit Quantum Transformer Block; mostly convincing proof of concept, but key circuit details are IP-withheld.
PINNForge: Execution-Grounded Evolutionary Design of Physics-Informed Neural Networks for PDE Solving via Large Language Models
Mingyang Yu, Xu Yang, Jun Zhang, Xiaolong Wang et al.
PINNsForge: LLM execution-feedback evolution wins 24/25 PDE tasks with solid stats; credible, but novelty vs Lang-PINN and uncounted LLM cost temper the claim.
WaveFront Decoding: Parallelized Self-Speculative Decoding for Looped Language Models
Hyeongju Ha, Jae-Joon Kim
Credible training-free co-batched draft-verify decoding for looped LMs with solid speedups; evaluation is narrow — two checkpoints, one GPU, greedy-only.
Enforcing Narrative Reliability and Epistemic Pacing in LLM-Driven Detective Games via Structured Knowledge Trees
Parsa Rahmati, Richard Zhao
Proposes SKT+tri-agent pipeline: critical hallucinations 17.80%→6.27%, premature confessions 10/33→0/33; promising but confounded evaluation, no code.
Anatomy of a Closed-Loop Collapse: A Causal Case Study of a Compressed VLA Policy
Fengze Jia
Claims a causal anatomy of one compression-induced closed-loop collapse: four therapies fail, teacher-rollout substitution suffices. Rigorous but single-run, n=1.
Optimizers for Diffusion Models: A Controlled Benchmark
Arman Bolatov, Egor Shulgin, David Li, Abduragim Shtanchaev et al.
First controlled optimizer benchmark for discrete diffusion; well-run and reproducible, but text-task budgets are just 3% of published training length.
From Concept Alignment to Causal Grounding: An Intervention Test of Chain-of-Thought Faithfulness
Qianli Wang, Yilong Wang, Dennis Wei, Jingyi Sun et al.
Reframes CoT faithfulness as SAE concept grounding with a validated causal metric Δp across 5 LLMs; strong validation, but conceptual and scope caveats remain.
Attributable Post-Rationalization in RAG Citations: A Controlled Reproduction and an RLVR Comparison
Mehedi Khan, Md. Shariful Islam Bhuyan
Claims answer-only RLVR leaves RAG citation faithfulness unchanged; controlled reproduction is persuasive but lacks human validation and an SFT control.
Bridging Static and Agentic RAG for Taiwanese Historical Question Answering
Kai-Hsin Chen, Wei-Yu Chen, Xuanjun Chen, Jyh-Shing Roger Jang
Agentic vs static RAG tie on aggregate yet differ on 70.8% of questions; post-hoc selection recovers 60% of oracle headroom — solid but LLM-judge dependent.
LazyAgent: Demand-Driven Materialization and Physical Optimization of Agentic Programs
Xin Heng
Paper proposes demand-driven backward-closure permission for agent programs, claiming 42–52% cuts vs stop-eager; evaluation is unusually rigorous but the regime itself is unproven.
Event Signature Transfer: Model-Agnostic Forecast Scenario Construction from Historical Events
Karthik Sridhar, Aaditya Jain, Murari Mandal, Saurabh Deshpande
EST is a training-free, quantile-only operator transferring past event effects onto forecasts; promising but entirely in-sample, with hand-set controls and no prospective test.
Directing large language models to follow the letter or spirit of the law
Peng Qian, Andrew Li, Sam Chen, Sonia K. Murthy et al.
Anchored-LoRA adapters steer LLMs to letter- or spirit-based judgment across vignettes, Reddit, caselaw; solid evidence, letter-side effects weaker.