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.
Rollback the World, Keep the Reflection: Rollback-Induced Reflection for Long-Horizon LLM Agents
Yi Yu, Liuyi Yao, Yaliang Li, Enshu Wang et al.
RIR unifies rollback and reflection for LLM agents; consistent gains on 3 benchmarks, but no significance tests and the optimality theory is a capacity claim.
Made in Hungary: Comments on the performance of generative language models
M\'aty\'as Osv\'ath, Enik\H{o} H\'eja, No\'emi Ligeti-Nagy
Hungarian LLM adaptations show contamination and capability loss; this critique is largely sound, though its Qwen3>Racka reversal partly stems from truncation.
SEA-LION-v4.8: A Technical Report
Ahmed Mohammad Dabeer (David Wang Dawei), Ahn Jeongmi (David Wang Dawei), Anocha Sutaveephamochanon (David Wang Dawei), Antonyrex Sajeban (David Wang Dawei) et al.
Adapts Nemotron 3 to SEA languages via CPT + online distillation with strong SEA-HELM gains (120B 49.30→63.44); solid engineering but evaluated only against its own base models.
Behavior2Value: Benchmarking and Empowering LLMs for Consumer Value Measurement from E-commerce Behaviors
Peixuan Hou, Bin Chen, Li He, Jian Xu et al.
New benchmark for consumer-value inference from e-commerce behavior; fine-tuned verification model beats prompted LLMs, but lacks significance tests and public data.
Exact semantic readout from compressed vector representations
Daniel Quigley
Clean rank criterion for when embeddings admit exact truth-conditional readouts; rigorous math, careful but narrow monadic transductive experiments.
Beyond Accuracy: How Procedural Traces Shift the Decision Criterion of LLM Overseers
Zihan Chen, Di Zhu, Lei Zheng, Weiling Li
Detailed traces leave LLM overseers' detection intact but shift criteria toward false rejections; well-controlled SDT study, limited to 19 visible-evidence items.
From a River in Gilead to the Inference Distributions of Large Language Models: Covert Dialect Bias and Linguistic Profiling at Scale
Chowdhury Mohammad Abdullah, Rita Orji
Extends covert LLM dialect-bias probing to Nigerian Englishes across ten models; consistent but small effects and LLM-generated stimuli temper the claims.
Visual Compliance via Executable Safety Rule Entailment
Jisoo Kim (Sungkyunkwan University), TaeYoon Kwack (Sungkyunkwan University), Jinwoo Jang (Sungkyunkwan University), Honguk Woo (Sungkyunkwan University)
GUARDEN turns safety rules into executable, scene-graph-grounded logic; impressive gains on its own benchmark, but external validity and statistical rigor remain unproven.
T-SANDHI: Tone Sandhi-aware Adaptive Network with Decoupled Hybrid Injection for Low-resource Taiwanese Hokkien Speech Recognition
Hung-Yang Sung, Chien-Chun Wang, Tien-Hong Lo, Yu-Sheng Tsao et al.
T-SANDHI claims foundation models handle Taiwanese Hokkien tone sandhi well but confuse sandhi with rare citation tones, and its dual-stream gated injection on frozen Whisper delivers consistent CER gains (up to 12.27% relative) — plausible and well-ablated, though significance testing, baseline fairness, and novelty claims need scrutiny.
Recursive Reasoning or Statistical Extrapolation? In-Context Learning in Multi-Agent Interdependent Decision-Making
Yu Liu, Wenwen Li, Yifan Dou, Guangnan Ye
ICL gains vanish when price trends break, favoring statistical extrapolation over reasoning; a REE-based diagnostic, but the reasoning null is not airtight.
Think Before You Comfort: Reflective Cognitive Alignment for Protocol-Grounded Elderly Stimulation Agents
Jiyue Jiang, Ziyi Li, He Hu, Sheng Wang et al.
RCA couples structured reasoning with inference-time selection for LLM CST agents; gains plausible but rest on weakly-agreeing LLM judges and synthetic data.
Hyperbolic Graph Representation Learning for Differential Diagnosis on Biomedical Knowledge Graphs
Pietro Miotto, Lucia Mellini, Tommaso Marzi, Cesare Alippi et al.
A self-described preliminary study claiming hyperbolic embeddings represent biomedical ontology hierarchies at much lower dimensionality and modestly improve patient-disease ranking, but with no clinical baselines, no significance testing, and a likely unfiltered ranking evaluation.
PrimeScientist: Strategic Allocation of Research Effort in Autonomous Research
Xinle Yu, Fan Bai, Kaiser Sun, Hengshuo Miao et al.
Budget-adaptive MCTS over executable plan trees allocates autonomous-research effort; attempt-efficiency gains look solid, but headline reward gains rest on single runs.
How Calibration Content Shapes Attention-Based Reranking
Petros Karypis, Hossein Rajaby Faghihi, Peter Chen, Rui Zhu et al.
Finds null-query calibration breaks on per-query instructions; training-free interpolated nulls fix it, though RankGPT still wins on FollowIR.
SFT or RL for Tool-Calling Agents? A Controlled Study Across Data, Method, and Scale
Md Tahmid Rahman Laskar, Xue-Yong Fu, Shashi Bhushan TN
A controlled SFT-vs-GRPO-vs-SFT→GRPO study across six Qwen3 scales and three tool-calling corpora concludes LoRA-SFT is the strongest in-distribution method and dataset choice matters as much as method; the design is unusually thorough, but near-one-point transfer margins lack significance testing and the eval metric exactly equals the RL reward, biasing the comparison toward GRPO—making the 'GRPO modestly better at transfer' claim soft.
Beyond Truncation: Rethinking LLM Decoding as Ensemble Pruning
Dunyao Xue, Chengshuo Du, Zhengbo Wang, Wenlin Dai et al.
ME-Decoding casts token selection as Mahalanobis ensemble pruning; temperature-robust gains, but near-greedy supports and vacuous theory temper the claims.
Where Should Agents Live? Energy-Memory Characterization of Agentic AI for the Edge-Cloud Continuum
Carolina Fortuna, Vid Han\v{z}el, Tim Strnad, Bla\v{z} Bertalani\v{c}
Extends eCAL to multi-agent LLM energy via a validated two-rate model; text hand-offs cost <0.25% of workflow energy, while multi-agent setups rarely pay off.
Infinite-Parameter LLMs: Generating and Adapting Weights from Live Data
Jinli Hu, Ross M. Clarke, Yichuan Zhang, Jos\'e Miguel Hern\'andez-Lobato
Generates LLM experts from live data with an online Bayesian belief over the code; categorical-router results look strong, but the Bayesian core is untested.
Beyond Routine Compliance: Cunning Data Cultivates Safety Vigilance in Large Language Models
Youjia Wang, Lin Xu, Yang Sun, Yuxiao Lu et al.
Claims Ruozhiba 'cunning' data transfers to LLM jailbreak robustness; interesting, but data-vs-method confounds and single-run results temper the SOTA claim.
Disentangling Long-Term Memory via Latent Neuro-Symbolic Reasoning
Cai Ke, Xinghao Chen, Xiaoyu Shen, Keyu Chen et al.
LGM's query-conditioned latent graphs show large gains on personalization benchmarks, but evaluation lacks significance tests, a stated judge, and code.
Reasoning through Evolution: Automatic Meta-path Discovery for LLM-based Fake News Detection
Ziyi Zhou, Xiaoming Zhang, Hui Pang, Yuting Zhang et al.
GA-evolved meta-paths let frozen LLMs use propagation graphs for fake news detection; strong low-resource gains, but no significance tests or variance.
Faithful yet Collusive: Why Chain-of-Thought Monitoring Cannot Detect Collusion in LLM Pricing Agents under Oligopolistic Competition
Dohun Lee, Hyunwoo Park
Clever causal-graph audit finds LLM collusion and CoT faithfulness dissociate across 9 models; evidence is suggestive but rests on very small samples.
Re2A: Situated Conversational Recommendation via Rubric-based Preference Reasoning and Alignment
Dongding Lin, Jian Wang, Xiaoyan Zhao, Wenjie Li
Rubric-guided preference reasoning plus preference-conditioned DPO for situated recommendation; strong gains (Hit@1 48.1, VHR 5.2%) with solid controls; judge-teacher overlap and 2x latency noted.
Market Signal Injection: Adversarial Context Manipulation of LLM Pricing Agents
Dohun Lee, Hyunwoo Park
Simulation study: presentation-only edits of rival inputs steer LLM pricing agents and markets; credible, but n=5 seeds and fixed demand limit scope.