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.
What Counts as Strategic Reasoning? A Systematic Mapping of Chess Research on Humans, Engines, and Language Models
Paolo Ciancarini, Remo Pareschi
A systematic map of 84 chess studies (2023–26) across humans, engines, RL, LLMs, and hybrids; valuable gaps, but single-coder coding limits reliability.
AeroWeaver: An Embodied-Agent Harness for Weaving Aerial Skills into Distributed, Adaptive Swarm Execution
Jiabin Lou, Yirong Yang, Haopeng Wang, Xuxin Lv et al.
An LLM harness weaving UAV skills, decentralized agents, and training-free reward memory; consistent simulated gains, but baselines and statistics are thin.
Bad Genius: Counterfactual-Guided Harness Evolution Beyond Task-Specific Shortcuts
Guojun Zhu, Xunheng Huang, Peng Yin, Jiahui Xie et al.
CHASE counters benchmark-wide shortcuts in LLM harness evolution via protocol counterfactuals; formalization strong, evidence thin (3 rounds, no repeats).
Cultural Competence in Context: A Large Language Model Passes the Turing Test in Finland
Otto Segersven, Pentti Henttonen
Claims ChatGPT 5.2 passes a Finnish-language Turing Test via model-generated role prompting; plausible but the 'pass' rests on non-significance with n=36.
HPOQuest: A Rare-Disease Diagnostic Agent Using Active Phenotype Acquisition
Kamilia Zaripova, Nassir Navab, Azade Farshad, Annalisa Marsico
Training-free sequential HPO questioning improves sparse rare-disease diagnosis—strong on the small MME cohort, mixed elsewhere, no significance tests reported.
WetRobo: A Reproducible Robot Kit for Coding Agents in Biological Laboratories
Yuna Oikawa, Kei Endo, Takanori Uzawa, Yunzhe Zhang et al.
WetRobo: a coding agent adapts a robot kit across two real wet labs and beats a fine-tuned VLA on cross-lab cap transfer; single-run, single-model evidence tempers the claim.
Building Trust in Artificial Intelligence: A Necessity for Railway Applications
Lefebvre Renard Cl\'ement, L\'eb\'e Vincent, Da Silva Ribeiro Pereira Ricardo, Sundell Johan et al.
Position paper mapping robustness, ODD and explainability onto railway safety standards; a solid bridge, but no experiments back its 'ready to use' claim.
Who Audits Whom, on What Substrate, with What Evidence? An Independence-Graded Audit Protocol for Agentic AI
Mohamed Chahine Ghanem
A three-axis independence grading scheme for AI audits, formally grounded and honestly caveated, but validated only by simulation with declared parameters.
BENCHCOMPASS: From Scores to Signals for Training and Harness Decisions in Payment-Domain LLMs
Sijie Dong, Wei Ren, Xuanwei Hu, Jiawei Luo et al.
Solid payment-domain LLM benchmark (306 items) separating knowledge, context-use, attack robustness; small subsets and judge-family overlap temper claims.
WFM: Wiki Foundation Model for Complex Agentic Reasoning
Junnan Dong, Linhao Luo, Senlei Zhang, Gong Chen et al.
A GFM over dense LLM-Wiki graphs claiming 10.5x speedup and SOTA QA gains; promising but a leaked 'planning values' note and absent stats weaken trust.
AutoTuneBench: Trustworthy Measurement for Agent Auto-Tuning of LLM Serving Engines
Li Chen
AutoTuneBench freezes a measurement protocol as code with architectural anti-cheat for LLM-agent engine tuning; convincing design and unusually honest negatives, but evidence is single-box and dominated by null results.
When Is Graph Structure Worth Its Cost? The Case for Structure Pricing in Retrieval-Augmented Generation
Yuzhong Zhang, Haoyang Ma, Chao Peng, Lionel Briand et al.
EffiRAG wins 93/120 judged queries vs LightRAG-hybrid at 57% lower cost; strong cost-aware protocol, but only a small LLM-generated benchmark backs the claims.
The Other Half of the Memory Wall: Serving 35B MoEs from SSD with Trained Routing Prediction
Yu Lin, Yiming Wang, Runyuan Cai, Hanze Liu et al.
Edge0 serves a 35B MoE from SSD at 20 tok/s inside 3 GiB via one-token-ahead routing prediction plus an unmerged recovery LoRA; solid, unusually honest systems work, but headline gains lack external offloading baselines.
Multimodal Conditioning of Fine-Tuned Stable Diffusion XL for Controllable and Culturally Faithful Ulos Motif Generation
Humasak Simanjuntak, Tamara Yunika Sianipar, Bronson T. M Siallagan, Difya Laurensya Ambarita et al.
Applies LoRA-tuned SDXL plus LLaVA and ControlNet conditioning to Ulos motif generation; its ablation finding that more conditioning is not better is interesting, but evidence rests on a 7-image test set with inconsistent metric reporting and no text-only baseline.
Missing Bridges: Composition-Aware Active Imitation Learning
Maxwell J. Jacobson, Ahmed H Qureshi, Yexiang Xue
Active IL: AALT requests bridge demos maximizing start-goal connectivity; clean theory, striking gains (100% with 3 demos), external validity untested.
TuiML: Machine Learning for AI Agents
Nilesh Verma, Nick Lim, Albert Bifet, Bernhard Pfahringer
TuiML: a self-describing ML library for LLM agents via MCP; coherent design, accuracy parity with scikit-learn, but the agent-benefit claim itself is untested.
RideWay: Benchmarking Efficient Task Completion for Tool-Using Language Agents
Qingnuan Han, Boli Fang, Mingzhi Hou, Claire Liu
RideWay adds a human-calibrated efficiency metric on top of agent task success; rigorous and candid, but the tool axis is chance-level and labels thin.
Decodability is Not Causality: Dissociating Probe Readouts from Behavioral Drivers via SAE Decomposition
Devesh Tiwari, Camille Davis, Shivank Sinha, Talia Weaver et al.
A workshop-scale but methodologically careful SAE-intervention study showing that probe weight geometry does not identify the Gemma-2-9B features that causally control a True/False truth probe's behavior — convincing within its single model/layer/task setting, but generalization is untested.
Learning Heterogeneous Preferences
Shiwali Mohan, Matt Hong, Dule Shu, Aniek Fransen et al.
Person-conditioned utility models beat universal ones on 575k wheel-aesthetic judgments; convincing ablation, but baselines, splits, and stats reporting thin.
ERPBench: A State-Grounded Evaluation Paradigm for Computer-Use Agents in Enterprise Software
Kratika Bhagtani, Kusha Sridhar, Maziyar Baran Pouyan, Yuying Zhao et al.
ERPBench grades screenshot-only agents by ERPNext database state, exposing silent wrong-value saves; strong idea, but 30 tasks, 5 runs, no significance tests.
Measuring AI Leadership: Development and Validation of a Multidimensional Measure for AI-Native Organizations
Mustafa Akben, Leslie Coyne
Develops a 108-item AI Leadership Battery with solid multi-stage validation; main concerns are an underpowered 51-item CFA and same-source criterion data.
Memory Has Geometry: Non-Uniform Geometric Memory for Long-Horizon Personalized AI
Jiahong Liu, Wenhao Yu, Zexuan Qiu, Menglin Yang et al.
Vision paper: long-horizon memory as trajectory-conditioned estimation on non-uniform geometry; compelling agenda, purely conceptual, no experiments yet.
SAGE: Governed Artifact Generation from Enterprise Guidelines
Mohammadreza Sediqin, Shivali Dalmia, Sumukha Thoppanahalli, Srinivasa Karthikeya Reddy Kovvuri et al.
SAGE cuts guideline-to-artifact work from days to ~1h at 3.2% hallucination via a governed LLM pipeline; sound design, but evaluation rests on proprietary data and proxy metrics.
When to Call an LLM: A Confidence-Gated Hybrid for Cost-Effective Emotion Recognition in Conversational AI
Sai Babu Udayagiri, Arjun Chouhan, Ravisekhar Kanagala, Trishala Pavagada
A credible study: confidence-gated ensemble+LLM hybrids beat both pure systems on three ERC benchmarks; mostly supported, though accuracy trails SOTA.