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.
Mo' Models, Mo' Problems: How to best select model pools when designing Multi-Agent Systems
Sara Vera Marjanović, Jiacheng Xu, Aleksandr Laptev, Grigor Nalbandyan et al.
Claims that growing heterogeneous model pools degrades MAS performance below its best single model; a systematic 23-model study, credible overall, but headline curves lack variance estimates.
Quantum-Inspired Trainable and Parameter-Efficient Tensor Networks for Image Inpainting
Shiwen An, Konstantinos Slavakis
Trainable QFT tensor networks ensure minimal coherence and strong parameter-efficient inpainting; single-seed evaluation and an omitted proof temper claims.
Rethinking Domain Specialization for Open-Ended Scientific Reasoning in Astronomy Language Models
Vanessa Lama, Sanjay Das, Emily Herron, Yuan-Sen Ting et al.
Argues frontier general LLMs beat astronomy-specialized models on open-ended olympiad astronomy QA; a sound multi-view evaluation, but headline conclusions rest on an unvalidated LLM judge.
Intrinsic Motivation in Reinforcement Learning: A Research Agenda for Adaptive Self-Organisation
Anatoly Belikov
A tutorial-style perspective arguing that intrinsic RL rewards (empowerment, DIAYN, curiosity, information gain) exhibit predictable failure modes curable only by environmental constraints and multi-agent communication; mathematically pedagogical and honest about limits, but entirely non-empirical.
Towards Detecting AI-Assisted Responses in Online Surveys
Qizhou Wang, Bogdan Mamaev, Christopher Leckie
Persona-grounded LLM agents evade existing survey detectors; SPABD's few-shot behavioural-cue aggregation restores +0.14 AUROC — credible but small-sample.
From Transient Prompts to Persistent Control: Scientific Poster Generation via Recursive Semantic-Geometric Contracts
Runze Li, Yukun Zhao, Can Xu, Yucheng Shen et al.
PosterVisor adds frozen rubric contracts and repair-aware rechecking to LLM poster generation; gains are real but modest, with ablation trade-offs.
Vroom-Vroom at SHROOM-Visions: A Multi-Judge Committee for Detecting Hallucinated Spans in Vision-Language Outputs
Toqeer Ehsan, Nico Penttilä, Richard Schmidt, Arash Hajikhani et al.
Five-model VLM committee wins 3 of 4 SHROOM-Visions languages; strong aggregation design, but vote threshold tuned on test split and no significance tests.
Self-Emergence Agent Architecture:Behavior-Inertia HMM, Reflexive Metacognition,and Social-Contrastive Self-Modeling
Xiaoyang Liu
Personality as an HMM matrix edited by verbal reflection: the mechanism prototype is causal, but the full LLM loop is never run end-to-end.
Type-IV Code Clone Detection via Layer-Wise Non-Contrastive Representation Learning
Luciano Marchezan, Kevin Delcourt, Eugene Syriani, Houari Sahraoui
Layer-wise non-contrastive VICReg beats contrastive and LLM baselines on Type-IV clone detection, but its own ablation favors a simpler last-layer variant.
A Spatiotemporal Extension of the Neuromorphic DBSCAN Implementation
Charles P. Rizzo, James S. Plank
Extends neuromorphic DBSCAN SNNs with causal temporal look-back via memory-neuron feedback; precise resource math and open code, but no efficacy results.
Coding Agents Have Converged: Why the SWE-bench Leaderboard Can No Longer Order Its Top Entries, and What to Measure Instead
Fengshuo Liu, Ying Liu, Ruize Sun, Lie Luo et al.
Auditing 254 SWE-bench submissions, the paper argues top-entry leaderboard gaps are statistically inseparable and scaffold-dependent - a rigorous, reproducible, mostly convincing evaluation audit.
Where Should a Document Live: Context, Representations, or Parameters?
Nathanaël Carraz Rakotonirina, Momchil Hardalov, Gonzalo Iglesias, Adrià de Gispert
Controlled study finds KV-cache Cartridges beat parametric knowledge injection in accuracy and composability; plausible but model-dependent (Gemma replication fails).
Bias-Induced Crossover in Absolute Capacity of Dense Associative Memory
Yuto Sakurai, Takeaki Shimokawa, Kazunori Iwata, Kazushi Mimura
Bias cuts dense-memory absolute capacity from N^(n-1)/ln N to O(N^(n/2)) near q=1/2; analysis is careful, but Gaussian-tail claims lack rigorous validation.
ECHO: A Matched-Contrast Benchmark for Context-Sensitive Turn-Taking in Full-Duplex Dialogue
Shuofeng Zhao, Hongwei Cai, Wenke Fan, Qingxiang Guo et al.
ECHO proposes a matched-insertion paired benchmark exposing YIELD bias in full-duplex turn-taking, but its central 'context-sensitivity' claim is weakened by label-dependent interruption acoustic processing and the absence of a gain-free or waveform-reuse control.
OPEN-1B: A Fully Auditable Training Run
John Donaghy, Brian Wilcox, Oğuzhan Ersoy, Shikhar Rastogi et al.
Claims the first bitwise-fully-auditable LLM training run (Open-1B) via RepOps, an invariant data stream, and deterministic collectives; engineering is credible and costs honestly reported, but the audit trust model rests on publisher-generated hashes.
JustFit: 200K-Token LLM Serving on a 24 GiB Laptop with Just-in-Time State Management
Yuhua Chen
Claims 6.93× context gain on a 24 GiB laptop via 4-bit KV + JIT state management; credible systems work, but long-context quality and ablations are thin.
Learning-Guided Planning in Large Dynamic Action Spaces: Budgeted Tree Search for One-to-Many Mobile Charging
Liang-Ching Tao, Pi-Chung Wang
Preregistered study of learned-prior budgeted tree search for sensor charging; rigor stands out, but the headline margin is statistically unresolved.
Right Tool, Right Job: Native-Language Evaluation, Tokenizer Sensitivity, and Methodological Findings from a French-Only BabyLM
Adam Zachary Wasserman, David Beauchemin
A French-only BabyLM model scores 85.97% on native QFrBLiMP with unusually honest ablations, but headline comparisons are cross-benchmark and several effects sit near its own noise floor.
CareMirror: Bringing Caregiver Wellbeing into the Dementia Care Ecosystem
Jiayue Melissa Shi, Ethan Nguyen, Drishti Goel, Upasana Natarajan et al.
Design-probe interviews (N=14) of CareMirror yield credible qualitative findings on caregiver-clinician data sharing, but lack clinician input and any sustained-use evidence.
Evaluating Verified Autonomy in Quantum Engineering
Naixu Guo, Changhao Li, Siyu Cheng, Qicheng Tang et al.
Evidence-bound quantum-agent benchmark exposes a 78%-to-3% reliability spread across 17 systems; strong design but simulation-only, n=3 trials, harness-confounded.
FreqSpaNet: Frequency and Spatial Learning of SFPF for Physical Layer Hardware Integrity Detection
Xiaoxuan Huang, Jinlong Xu, YiZhe Wang, Meng Zhang et al.
Dual-branch SFPF network detects wireless hardware replacement (96.31% AUROC); strong design and ablations, but lab-only data, no code, no significance tests.
Graphlets as structural fingerprints of complex networks
Anna Pidnebesna, David Hartman, Aneta Pokorna, Daniel Trlifaj et al.
Graphlets beat classical graph metrics on synthetic and rewired networks but only match them in schizophrenia classification — a careful, convincing benchmark.
LACE: Layer-Wise Compression for Dynamic Frame Rate Codecs
Thanapat Trachu, Samuele Cornell, William Chen, Shinji Watanabe
LACE claims per-layer segmentation in dynamic frame rate codecs yields a better rate-quality tradeoff, and largely delivers: rigorous theory and controlled reconstruction gains, though TTS evidence is narrow and its efficiency framing is asymmetric.
Coupled Calibration and Learning: Mitigating Teacher Bias in LLM Distillation without Target-Domain Reward Feedback
Haichen Hu, Yuheng Zhang, David Simchi-Levi
CCL provably distills an unbiased oracle student from a biased teacher using only source-reward feedback; rigorous but assumption-heavy theory, no experiments.